Best Practices Series
Knowledge. Yours Free. In Every Issue.
Hear directly from the vendors on a variety of topics designed to help you make better decisions. Round Table: Each series includes an accompanying round-table webinar where you can learn more about each topic and have your questions answered by industry experts. Get the schedule at www.speechtechmag.com/webinars.

July/August Issue
• Speech Is an Essential Business Tool • Engaging Customers in Mobile Environments • ROI in Speech—Invest in New Technology or Leverage Existing Infrastructure Reservations Due 5/20/11; Copy Due 5/23/11

Sept/Oct 2011 Issue:
• Speech Analytics • Speech in Government • Speech Targeting Senior Audiences— How to Best Utilize and Gain Acceptance • Secrets to Better Customer Service Using Speech Reservations Due 7/22/11; Copy Due 7/25/11

Nov/Dec 2011 Issue:
• Hosted Solutions • Speech Automated Services on Mobile Phones Reservations Due 9/16/11; Copy Due 9/22/11

Vendors: If you’d like to participate in one of these special supplements, please contact an Integrated Marketing Manager: Adrienne Snyder Eastern & Central 201.327.2773 Dennis Sullivan Mountain & Pacific 800.248.8466 x. 538



> > F E AT U R E S

May/June 2011
Volume 16, Number 3


Speech Analytics Captures Consumer Sentiment
Technologies blend data from different sources, emotional states and moments in time.


Hosted IVRs Bring Peace of Mind
Price and flexibility are encouraging corporations to switch from premisesbased systems.


Don’t Let M&As Steer You Off Course
Tech vendors gets acquired. What can your organization do to protect its investments?

>> FYI


10 11 12 12 13 14 15 34 36

SpeechCycle Dedicates Division to Smartphone Customers
Interactions over mobile devices are so different that SpeechCycle created a unit devoted to customer care. The first application, Smart Care Mobile, takes calls out of the contact center to optimize the customer experience.

2 4 8 9 38 39 40

Editor’s Letter
Making a Case for IVR Hosting

In Other Words
Global Presentations Using TTS And Automated Translation

Call Centers Receive Guidance on Protecting Credit Card Data Based
on the Payment Card Industry’s Data Security Standard, service providers get recommendations to ensure the integrity of phone-based transactions.

Tech Investments Buoy Workforce Optimization The WFO suite
market is forecast to post healthy growth during the next few years.

Inside Outsourcing
Spring-Cleaning or a New House?

Overheard/Underheard Under-the-radar speech news. Mobile: The New Frontier for Voice Security Nearly a third of
consumers believe voice biometrics is the best method for verifying their identities on mobile devices.

Much Ado About Toronto

Market Spotlight: Government Homeland security drives voice R&D. Soundbytes News roundup.

Making Modalities Play Nicely Together

IVR Is Good News for Scripps The media company leverages Pronexus
system to standardize subscription services for 14 newspapers.

Voice Value
Emergency Preparedness Is a Job for All

Forward Thinking
If This Passes, Only Criminals Could Do Speech Technology

Rural Arizona Connects With Residents Switching from a hosted to an
on-premises contact center suite speeds service and cuts costs. BY LEONARD KLIE



Speech Technology | 1


Editorial Director David Myron dmyron@infotoday.com News Editor Leonard Klie lklie@infotoday.com Managing Editor Greg Lupion glupion@infotoday.com

Making a Case for IVR Hosting
uilding and maintaining an IVR on premises can be complex and costly. In addition to the costs of deployment, periodic upgrades should be expected. But the details of the maintenance and upgrade costs raise questions: When is the next upgrade? How disruptive will the deployment be to systems and processes? How long will the upgrade take? How much it will cost? There are situations—as suggested by our cover story, “Hosted IVRs Bring Peace of Mind” (page 16)—in which on-premises IVRs are more economical than hosted IVRs. However, on-premises solutions require a large up-front capital expense. And, unless an organization sets aside money and resources to maintain and upgrade its IVR system periodically, it can easily get stuck with antiquated technology—especially during prolonged economic slumps. According to research cited in our story, the on-premises IVR market has been in decline in recent years during The Great Recession. Conversely, the hosted IVR market has been trending upward. Many who still resist the hosted model argue that organizations lose too much control of valuable data, processes, and customer interactions. And, unfortunately, some fool themselves into thinking they can do a better job than a company that specializes in customer service. That is akin to saying, “I don’t want to lose control, so I’m going to fly the plane instead of an experienced pilot.” Additionally, farming out the IVR to a third-party vendor doesn’t mean you have to lose control, either. Too many organizations mistakenly think that handing an important business process to a vendor means that they should throw the project or process over the fence and walk away. On the contrary, organizations should get regular reports and necessary alerts from the hosting provider, and they should periodically discuss opportunities and concerns that undoubtedly will arise. Those who recognize the value of the hosted model appreciate the fixed and predictable costs. Plus, they get technology upgrades at the push of a button at no additional cost. They also get a scalable system, so when IVR usage dips, the client can scale down to cut costs and scale up as needed. If that doesn’t convince you to consider a hosted IVR, contemplate this: What happens if your IVR vendor gets acquired? Acquisitions are common, according to the feature, “Don’t Let M&As Steer You Off Course” (page 28). It states, “In an industry rapidly evolving, acquisitions have become a common way for vendors to expand their businesses. Last year brought more than a dozen purchases…and Nuance Communications Inc. has made more than 30 acquisitions—including nine in 2007—since the turn of the millennium.” This article offers tips for navigating a new course if your on-premises speech vendor gets acquired. On the other hand, if your hosted IVR solution provider gets acquired, it’s immensely easier to switch to another hosting provider than it is to try to protect your on-premises investments. The benefits of the hosted model are clear, but if you’re still not sure, then conduct a low-cost, no-risk test run—something you can’t do with an on-premises solution. But, first, read our cover story on the topic.

Senior Designer Laura Hegyi Proofreader Greg Edmondson Contributors Kevin Brown, Deborah Dahl, Paul Hyman, Paul Korzeniowski, Melanie Polkosky, Sue Ellen Reager, Robin Springer, Moshe Yudkowsky
Editorial Advisory Board The Editorial Advisory Board of Speech Technology magazine is composed of prominent figures in research, development, and applications of speech technology. The members will assist the magazine’s editorial staff by recommending articles and column topics or prospective authors, and offering advice on technical matters and industry trends. Additional responsibilities will include critiquing the magazine’s content and design. If you are interested in participating, contact David Myron, Editorial Director, Speech Technology magazine, 237 W. 35th St., 14th Floor, New York, NY 10001, (212) 251-0608.


ADVERTISING SALES / PRINT & ONLINE Eastern & Central Ad Director Adrienne Snyder adrienne@infotoday.com (201) 327-2773 • Mountain & Pacific Ad Director Dennis Sullivan

dennis@infotoday.com (800) 248-8466, ext. 538
MARKETING Marketing Manager, Events & Circulation Sheila Willison sheila@infotoday.com • Director of Web Events DawnEl Harris dawnel@infotoday.com CORPORATE HEADQUARTERS

Information Today, Inc., 143 Old Marlton Pike, Medford, NJ 08055
EXECUTIVE MANAGEMENT President and CEO Thomas H. Hogan • Chairman Roger R. Bilboul • Vice President, Administration John Yersak • Vice President, Content Dick Kaser • Group Publisher Bob Fernekees INFORMATION TECHNOLOGY Vice President, Information Technology

Bill Spence

PRODUCTION Vice President, Graphics and Production M. Heide Dengler • Ad Trafficking Coordinator Michael Hardwick

Speech Technology (ISSN: 1088-5803) is published six times a year (January/February, March/April, May/June, July/August, September/October and November/December) by Information Today, Inc., 143 Old Marlton Pike, Medford, NJ 08055 USA; Phone: (609) 654-6266; Fax: (609) 654-6266; Internet: www.infotoday.com. Registered in U.S. Patent & Trademark Office.
Periodicals postage paid at Medford, N.J., and additional mailing offices. ©, Copyright, 2011, Information Today, Inc. All rights reserved. No part of this publication may be reproduced in whole or in part in any medium without the express permission of the publisher. P R I N T E D I N U S A POSTMASTER: Send address changes to Speech Technology, P.O. Box 3599, Northbrook, IL 60065 Rights and Permissions Permission to photocopy items is granted by Information Today, Inc. provided that a base fee of $3.50 plus $0.50 per page is paid directly to Copyright Clearance Center (CCC), or provided that your organization maintains an appropriate license with CCC. Visit copyright.com to obtain permission to use these materials in academic coursepacks or for library reserves, interlibrary loans, document delivery services, or as classroom handouts; for permission to send copies via email or post copies on a corporate intranet or extranet; or for permission to republish materials in books, textbooks, and newsletters. Contact CCC at 222 Rosewood Drive, Danvers, MA 01923; (978) 750-8400; Fax: (978) 646-8600; www.copyright.com. If you live outside the USA, request permission from your local Reproduction Rights Organization. (For a list of international agencies, consult www.ifrro.org.) For all other requests, including making copies for use as commercial reprints or for other sales, marketing, promotional and publicity uses, contact the publisher in advance of using the material. For a copy of our Rights and Permissions Request form, contact Lauree Padgett, lpadgett@infotoday.com. Online Access Visit our Web site at www.speechtechmag.com Searchable archive of all articles with digital document delivery: www.iti-infocentral.com Contents also available online under direct licensing arrangements with EBSCO, NewsBank, ProQuest, Gale, and H.W. Wilson and through redistribution arrangements with information service providers, including Dow Jones Factiva, LexisNexis, OCLC, STN International, and Westlaw. Subscription Information SUBSCRIPTIONS: Print version free to qualified recipients within the U.S. Digital version free to qualified recipients worldwide. Subscription rates for nonqualified subscribers: U.S. subscription rate—$29; Canada and Mexico—$94; overseas delivery—$130. All rates to be prepaid in U.S. funds. Subscribe online: www.speechtechmag.com/subscribe. BACK ISSUES: $8 per copy, prepaid only. CHANGE OF ADDRESS: Mail requests, including a copy of the current address label from a recent issue and indicating the new address, to Speech Technology Magazine, P.O. Box 3599, Northbrook, IL 60065-9736 or call (847) 559-7301. Reprints For quality reprints of 500 copies or more, contact Sheila Willison at (877) 993-9767, sheila@infotoday.com.

David Myron Editorial Director dmyron@infotoday.com @dmyron on Twitter

Disclaimers Acceptance of an advertisement does not imply an endorsement by the publisher. Views expressed by authors and other contributors are entirely their own and do not necessarily reflect the views of the publisher. While best efforts to ensure editorial accuracy of the content are exercised, publisher assumes no liability for any information contained in this publication. The publisher can accept no responsibility for the return of unsolicited manuscripts or the loss of photos. Privacy Policy Occasionally we make a portion of our mailing list available to organizations whose products or services we think might be of interest to our customers. If you do not wish to receive such mailings, please send a copy of your mailing label with a request to be removed from the third-party mailing list to Speech Technology magazine, Customer Service, P.O. Box 3599, Northbrook, IL 60065-9736 or call (800) 248-0588 or (847) 559-7301. Editorial Office 237 W. 35th St., 14th Floor, New York, NY 10001; (212) 251-0608; www.speechtechmag.com

2 | Speech Technology


DELIVER CARE WITH CONFIDENCE Customer service excellence is positively linked to retention if automation is implemented correctly. Integrated datadriven solutions provide the following benefits: • Complete calls quicker with dynamic IVR menus that recognize situational realities – For example.com/OnDemand or call 866-436-1169. see and share caller experiences. • Identify up sell or cross sell opportunities – Knowing a customer is a new mover may provide a bank an opportunity to offer home equity loans. Larger and more interactive dialogs now cost effectively available with many speech recognition systems are working to eradicate old perceptions of poor performance initially established years ago. Call recording cost effectively captures the more subjective aspects of customer satisfaction. A more effective approach is to also analyze the “voice of the customer” across all channels of communication. Automated dialogs allow customers to be more frequently understood the first time (even complex dialogs) without repeated requests. new account welcome messages.” How does any organization really know beyond simple statistics if their moments are diamonds or VoltDelta O OnDEMAND SOLUTIONS R duds. EVALUATE “VOICE OF THE CUSTOMER” ACROSS ALL CHANNELS Every interaction is potentially a customer care “moment of truth. examine the features available in leading voice self service systems to achieve this goal. Make sure they include: DATA-DRIVEN PERSONALIZED AUTOMATION Using internal and/or external data to deliver personalized automated speech allows organizations to know who is calling and predict why they are calling. Before investing in a solution. but at the expense of even more unhappy callers? How can financial institutions improve customer perceptions and contain service costs? The answer is to make it easy and convenient for customers to do business with you using a capable hosted speech self service solution. but do not forget the importance of entrusting customer care to systems that are reliable. VoltDelta’s patent-pending CrystaWAVE technology uses multiple simultaneous grammars with context sensitivity to: • Recognize with higher accuracy. • Disambiguate between like sounding words with greater confidence. creating distribution lists and scheduling campaigns. For example. Changes in regulatory disclosure rules prompted one VoltDelta customer to use the DeltaCast service to notify their clients of account inactivity. Robust hosted outbound systems should also include the ability to reach out and then convert responses with an inbound speech recognition application or agent support to deliver immediacy with cost savings. Organizations around the world partner with VoltDelta to deploy contact centers and home agents. enable self service with patent-pending speech technology.voltdelta. one financial institution recently switched to the VoltDelta platform due to a proven ability to handle millions of call per month with established infrastructure to successfully manage spikes in call volume. voice self service can now be a competitive differentiator delivering superior customer centric services at a lower transaction cost than a live agent by using a hosted model. and notification of suspicious account activity. More information on VoltDelta OnDemand can be found at www. ABOUT VOLTDELTA ONDEMAND VoltDelta OnDemand is a global provider of voice recognition automation and virtual contact center solutions. • Providing agent screen recording to identify the agent’s perspective. and communicate with multi-channel support. An organization can get a more thorough impression of the caller’s experience by: • Recording the experience as a single entity through any transfer. Over 2 Billion calls per year are processed within VoltDelta’s on-demand infrastructure. Real-time reporting provided managers with instant insight into the effectiveness of each outbound campaign. especially when a customer is transferred between agents and/or channels of communication? Statistical analysis alone cannot paint a complete customer care picture. such as a SMS text confirmation.Sponsored Content | May/June 2011 Speech In Finance Doing the Right Math: How Financial Institutions Can Win Back Customer Confidence While Controlling Costs Improve customer satisfaction AND reduce costs! This is the “post economic crisis” battle cry of senior leadership within many financial institutions that triggers despair for customer care managers.000 automated calls were made without IT involvement via an easy web-based configuration interface for uploading files.com . Over 30. • Remembering customer preferences expressed in previous interactions. Effective outbound campaigns include credit card activation notices. Speech Technology Magazine | www.speechtechmag. • Reduce deployment and ongoing tuning requirements for persistent cost savings. IMPLEMENT HIGH QUALITY VOICE RECOGNITION Advances in voice recognition technology now accentuate efficiency AND effectiveness. Beyond the obvious benefits of providing 24/7 access for many interactions and quick call resolution (if implemented correctly). PROACTIVE OUTREACH REDUCES INBOUND INQUIRIES Outbound calls help to improve customer satisfaction as well as reduce inbound calls. immediately prompting for a balance check to a repeat caller. • Providing a graphical user interface to easily listen. Aren’t adding agents and cost savings mutually exclusive? Do “Press or Say 1” voice self service systems save money. For example.

check boxes. and handle the account. desktop. easier. the page will swap for “nextpage. in order to compensate for the latency associated with the global online world of text-to-speech. The combination of TTS and automated translation repBuilt into PowerPoint is the capability to be a stand. get audience. TTS and automated translation reverse the impediment of language. for international presentations. But what is far less impactful than both is not izing their presentations. and automated translation movies. ity. voice recording. She can be reached at sueellen@internationalservices. the technol. in addition to refreshing stale content. and cheaper.mated translation less impactful than a professional transing issues—discourages 99 percent of companies from global. provide an online interface for to-speech can be improved with music in the background. worldwide customer support. Like TTS.lator? Yes.user’s access to the Internet. Web slides. and its clarifies the content.SUE ELLEN REAGER IN OTHER WORDS Global Presentations Using TTS And Automated Translation It’s the future of sales materials. Slides rotate as audio automated translation is in its adolescent stage. after 12 seconds. especially in other languages. PowerPoint is mechanisms add spice and interest to a page while providpopular for its ease of use.html” /> Making a difference As a technology. Fields. and pages. T TTS with Web presentations The concept of a TTS audio file associated with a slide also applies to Web pages by boosting multilingual online help and how-to instructions. The Web brings the freedom to modify and expand content more easily. demos. Is autoorganize the effort. vendors using online software.Translate 10 or 20 languages. and global system testing for speech and touch-tone applications as well as media localization. with the audio launch timing defined results cannot be used “as is. The program functions on all ing valuable feedback to its creator. and enable others to improve their customer support worldwide. ensemble creates a revolving presentaand improved by in-house personnel or by Automated translation tion for DVD. adding languages and new TTS. the results of the texttranslation.com. computers and all languages and is versatile enough to be saved in other formats.url=nextpage. polished. Then. TTS may buffering and playing online audio. users to modify content. Importantly. but at this adolescent stage. Examples and TTS enable audio Translations can be used to generate of the latter include http://www. a translation and software solutions company that performs translation.speechtechmag. 4 | Speech Technology MAY/JUNE 2011 www. as well as JPEGs. Though recording a tations to be translated into 10 or 20 languages at a lower human voice talent may make the greatest impact on the tar. install the audio. global cusalone. need for international voice talent YourWebpage. ogy can meet most industries’ needs by providing vital sales Web page presentations with audio also offer interactivand customer support.resents the future of international sales materials. and the cost of human record. replacing the YourVideo.corrections or modifications are faster. with audio acting as the presenter. TTS.” But the results of automated in the custom animation settings. audio with PowerPoint is a fine solution lations to improve automated translation results. the timing of a Web page swap should be longer than the exact length of the associated TTS audio Sue Ellen Reager is CEO of @International Services. Web pages can be timed to move from one page to the next by inserting a meta tag in the <head> of a page. The following code dictates that. The audio can be tomer support. and worldwide corporate media. TTS. ing—plus the management and administration required to Is TTS less impactful than a voice talent? Yes. including Web presentations. and takeAutomated translation and TTS enable audio and presenaways from global meetings or shows.html”: <meta http-equiv=“refresh” content “12. or the Web. which store autorecording. and predetermine preferred transWhen translated. TTS is powerful. and other response Among marketing and sales professionals. Nonetheless.com.Subtitle and presentations to be translated into TTS in other languages. as dictated by the not yet be emotive. and corporate media he most exciting—and underused—technology is file.com and http://www.cost with less administrative management. TTS offers an affordable solution.com . which is the current status of the industry. Together. translating at all. TTS could help thousands of companies increase their international support and help small companies go global. multiple choice. reduce the cost of pursuing international sales for companies that never could have afforded globalization. The translation can be processed. ROI.

August 8–10. 2011 NYC Will You Be There? .

Mobility is a game-changer and SpeechTEK is the premier venue for interacting with speech industry experts. C O M .AUGUST 8–10. and peers. 2011 HILTON NEW YORK.com Rates go up July 8! Gold Sponsor: Bronze Sponsor: Media Sponsors: Connect with SpeechTEK #SpeechTEK Organized and produced by Smartphone Interactive Users SCAN HERE W W W . This makes SpeechTEK your best opportunity in 2011 to gain the critical skills and solutions you’ll need to connect with the mobile customer. NEW YORK CITY CONNECTING WITH CUSTOMERS The integration of smartphones into customers’ daily lives is creating an ever-greater expectation of always-available customer service across multiple channels. vendors.SpeechTEK. S P E E C H T E K . Money-Saving Tip Use this code to save! STM11 Register online today: www.

industry consolidation. Plan and manage successful speech deployments. An organization must enable customers to get the same information via any channel (IVR.WHO SHOULD ATTEND? C-level executives Enterprise IT decisionmakers Telecom IT decisionmakers Customer service managers Customer experience designers Contact center professionals Speech solutions managers and developers Marketing executives Voice user interface designers Consumer electronics designers Anyone who wants to learn about deploying speech technology for business applications K EY N OT E S MONDAY 9:00 a. Strategic Growth and Planning. The days of separate voice-only and visual-only interactions are gone. KEYNOTE PANEL Mobility — A Game-Changer for Speech? Bruce Pollock. ATTEND SPEECHTEK 2011 TO LEARN HOW TO: Meet business goals with speech. and describes how voice user interfaces will be integrated with graphical user interfaces. what’s needed to move the speech industry forward? Speech technology holds the promise of enabling customers and businesses to connect in a timely. Increase adoption and acceptance of self-service offerings. Select the best technology for your business problem. discusses how speech technologies will be used by these applications.m. and customer distaste for automated IVR systems. can inspire a severe backlash that threatens a change in leadership and government stability. presidents The internet. web. mobile applications. mobile devices. Executive Director of Technical Research. – 10:00 a. as recent events have shown.m. Vice President.m. W W W . Use the latest advances in speech technology. WELCOME & OPENING KEYNOTE Responding to the Voice of the Constituent/Customer David Gergen.m. and social media are having such a profound effect on the way people acquire and share information. This panel of industry experts describes what new applications will be supported on mobile devices. or face-to-face interaction) and enable them to switch channels seamlessly without losing information. TUESDAY 9:00 a. Diagnose and solve speech application problems. and engaging manner. Will users embrace voice as they have embraced keypads on mobile devices? Are speech recognition and natural language processing able to process user speech into mobile devices? Will speech-enabled mobile applications replace IVR applications? Learn the answers to these questions and more during this keynote panel. – 3:15 p.m. Apply the latest principles in voice interaction design. Drawing on his extensive political experience and acumen. Failing to do so. Choose the right tools and environments. Senior Political Analyst for CNN. Program Chairs. AT&T Vlad Sejnoha.m. Former Adviser to four U. CLOSING KEYNOTE Prescription: What’s Needed to Move Speech Forward? DON’T MISS! SPEECHTEK LABS SPEECHTEK UNIVERSITY Check ou t the enclosed progra for det ail m s Susan Hura and James Larson. but how can businesses achieve this goal? The hard reality is that the industry must embrace change or become obsolete. From the latest U. Design a successful speech customer experience. presidential and congressional shifts to political unrest in the Middle East.S. S P E E C H T E K .S. Senior Vice President and Chief Technology Officer. WEDNESDAY 2:15 p. Co-Founder & CTO. – 10:00 a. This closing keynote reviews the best ideas presented at the conference and adds suggestions for how to improve the interaction between customers and businesses in this smartphone age. Improve customer satisfaction and loyalty. an inescapable truth has emerged: People want to be heard and they want the governments and organizations that serve them to respond to their needs. David Gergen will explain why this is happening and how businesses can avoid the mistakes of government leaders. Nuance Communications New mobile devices are dramatically changing how customers interact with businesses. Vlingo Mazin Gilbert. effective. social networks. SpeechTEK Conference SPEECHTEK EXHIBIT HALL With IVR market saturation. C O M . West Interactive Mike Phillips.

of Enterprise Integration Group (EIG). you would quickly conorate foyer? firm whether your belief is correct. and how important is it to solve it?” Moreover. kevin. and he can be reached at Bruce Balentine. applications in that you have added to the house. you feel good but you still have the same home standing the problem the IVR will solve. requirements better placed in the broom closet of your new. where he requirements. this activity and has a good flow throughout. pail. in through the less formal side door? Why not require many changes before you outdo some of them listen to that annoying man source your applications—you might conwho posted a map on the Internet that shows people how sider having a VUI designer perform an audit to provide an to get inside your house without going through your elab.external perspective.clearly provide examples that you may have never heard ing a new well-designed home. is much like spring-cleaning in your house: When you are The responsibilities of VUI designers require underfinished. However. you house that should serve you well for years to come. but they all don’t play ager who stuck around for less than a year) and the various identical roles. enced VUI team.speechtechmag. if you are building a new home. from both the with which you started. where all guests hate having you force a VUI designer is the correct choice for speech-enabled them through it.c. The speech technology industry isn’t as “travel rooms” featuring different languages that do not well defined as others. to set the proper application will help you sketch out a high-level perspective of your budget expectations. specializes in speech solutions design. the “really useful” addition to the back of the house Customer experience designers and interaction designdesigned by your eccentric uncle (the IT telephony man. 8 | Speech Technology MAY/JUNE 2011 www. Furthermore. want to ensure that you are fulfilling your current Kevin Brown is an architect at Hewlett-Packard Enterprise Services. A dust mop. application. uses the term “never asked question”: “What problem are W most of them review their existing applications to hen organizations consider outsourcing their IVR. hypothesis. Seldom are foyer. unless you called into your IVR regularly as you spring-cleaning of your house. For starters. hence the confusion over small match the rest of the house.KEVIN BROWN INSIDE OUTSOURCING Spring-Cleaning or a New House? VUI designers can help sort out requirements for outsourcing your IVR we solving.brown@hp. to the If you believe you own a house that is good shape point where they should quit asking to come already in good order—meaning it would before they are outsourced. and broom would be Why call them “true requirements”? Well. rather than a daylong before. Shouldn’t they feel grateful these activities. degrees of difference. well-designed come in many forms. If you agree with that made changes over time. as opposed to imagined or outdated ones. do yourself and your organization a new home? big favor.com .ers are similar to VUI designers. designers must maintain a conthis is the house that now resembles something a 7-year-old sistent theme across multiple applications in an IVR (think child might build out of Lego bricks. So. He has 18 years of experience in designing and delivering speech-enabled solutions. By doing so. Don’t call a cleaning crew until you have thorVoice user interface (VUI) designers are excellent oughly reviewed your existing applications with an experiresources to help you define your true requirements. you have decor in your house). And the audit would Perhaps outsourcing your IVR should be more like build. caller’s perspective and the goal of the owner of the IVR With additions and major renovations over the years.com. you want to ensure that your house is welcoming seek opportunities for improvement. Given that the discusDon’t forget the confusing and unfriendly sion centers on speech-enabled applications. then to whom should you turn for help in Seldom are speech-enabled applications in good shape determining what your real requirements are? And who before they are outsourced.

based interactions.MELANIE POLKOSKY INTERACT Much Ado About Toronto Watson’s Jeopardy gaffe underscores importance of resolving miscommunication nless you’ve been living under a moss-strewn rock. to Watson’s acceptance. just of a lack of information. then I think the answer is Chicago. When we can’t assume an easy cause. For example. That may be fine language usage for a TV game show Trebek: And you’d be correct. I think the answer might be Toronto.using that method. She can be reached at polkosky@comcast. in Toronto or anywhere else. knowledge. resort to emphasis of specific points. Watson causes are perceived as being the responsibility of both conresponded to a clue about a U.S. Model failures are more complex and occur because errors can be resolved outside the awareness of users. U www.a single turn of conversation. we puting. son wasn’t designed to converse but simply to answer and Trebek: [incredulous] Seriously? Toronto’s in Canada. chuckling) Didn’t you but not so much in the practical world of real interaction. or The communicative constraints of the Jeopardy game even three-quarters. because they require the least cognitive effort and are pathologist who has researched and designed speech. Interestingly. Trebek. content. They aren’t something to avoid and debug at all fidence in his Toronto answer. analogy. And that isn’t trivial tants. when miscommunication is viewed as tion (lexical failure). if we assume a perceptual and senior consultant at IBM. and other cognitive issues.com MAY/JUNE 2011 Speech Technology | 9 . Yet a Final Jeopardy while lower-level failures are generally seen as the responmistake on the first day had tongues wagging and fingers sibility of the speaker to fix independently. For three days in February. a socially skilled human Watson’s Toronto pbs. this miscommunication sage. While the surprising error in an Watson’s Toronto breakdown was a model failure otherwise dominating performance has because of inadequate background been analyzed ad nauseam (http://www. guys bother to feed Watson a map or a geography book? Most designers know writing error recovery is half. you’re aware that IBM recently either pulled off a However. is a social-cognitive psychologist and speech language chy. of the real work in dialogue design. And. such errors are resolve that by expanding his knowlmodel failure because of inadequate background constants in everyday human interacedge base. inadequate inferencing.as they are in our daily interactions. We had the correct one as another low-confidence response. itself worked against Watson because there is no option for in contrast to our engineering colleagues. they are a natural part of everyday conversation. and repeat the message verbatim. and multimedia user experiences for almost 15 years. Ph. the fix is generally to increase vocal intensity. IBM’s intelligent speaking super.net. we’re pretty cool resolving problems through dialogue. The two broad categories of can successfully shift from the highly constrained Jeopardy conversational error are input failures and model failures. and. airport with the answer versational partners to resolve. incorrect or no interpretation of informa.speechtechmag. and other explanations to get the concomputer Watson not only played Jeopardy but also handily versation back on track to shared meaning. She is currently a human factors psychologist the easiest to resolve. costs. ply speaking to Mr. Theorists suggest that when we recognize a problem during conversation. examples. In don’t gasp and stay up all night if someone gets an error mes. failure). As in other language(perceptual failure). As advanced speech technology moves from research Same as when we speak to people. depending on the blogosphere pundit in your earshot. graphics. we assume the lowest levels of the hierarMelanie Polkosky. followed by much explanation of the gaffe. or misunderstood meaning (syntactic inherent and designed for with an eye to human behavior. we make sure we correctly anticipated the could have been painlessly resolved. even more importantly. “What is Toronto?” Oops. Watson had low conwith errors. select clues.D.org/wgbh/nova/tech/ferrucciconversational partner would quickly breakdown was a smartest-machine. Of course. instead of I’ve been known to argue passionately playing Jeopardy: that miscommunication handling is the most important part Watson: Well. different belief systems of interac. ardy. Stuff like that happens cause of the breakdown. experiment to practical application in business. (aside. and the user successfully advanced the dialogue. beat two celebrated human contestants. exchange to a more robust interaction that allows for easy The former can stem from an inability to hear information resolution of miscommunication.. higher error types would not be quickly resolved massive marketing stunt or ushered in the next era of com. tion and really no big deal.inference explication. failure first. its usefulResearch on breakdown reveals a hierarchy of error types ness and usability largely will be determined by whether it that occur in every conversation. Instead. offered the appropriate message to us humans all the time. which slows the rate. higher-level pointing. but I’m of building intelligent speaking applications.html). Imagine if Watson was simknowledge. in the highly constrained interactive style of JeopWatson: OK. Watnot quite sure.

they are accustomed to watching videos.com “ ” .> > N E W S > > T R E N D S > > A N A LY S I S SpeechCycle Dedicates Division to Smartphone Customers SpeechCycle has launched a mobile division that will focus exclusively on meeting the growing demand for customer care solutions on the smartphone.speechtechmag. Google’s Android. “The paradigm of customer interaction is different. and increases revenue. It creates a highly It allows customers to bypass traditional IVR menus and get right to the applications they want to use.” he adds. SmartCare Mobile. which offers significant opportunity for companies to deliver a highly differentiated customer experience that drives customer loyalty. and RIM’s BlackBerry. which includes the following key features: systems—including Android. Shortly thereafter. Several corporations are piloting the application. The Mobile Division of SpeechCycle will dedicate significant resources to creating products and solutions that support the operating systems of leading smartphones and other mobile devices.” “Today’s consumers have become power users of devices like the Android and iPhone. “The telephone as we know it is changing into a much more powerful device. “It allows customers to bypass traditional IVR menus and get right to the applications they want to use. and since it has a builtin display.” Pan says. The need for the mobile division emerged when the company began looking at the smartphone market. chief technology officer at SpeechCycle. and the technology is different. Companies can plug SmartCare Mobile into their existing customer service applications to create custom widgets for customer surveys and feedback. and • seamless integration with enterprise contact centers and backend systems. and www. and scheduling. The top three right now are Apple’s iPhone. support.” says Alan Pan. • intuitive navigation with T3 (touch/type/talk) interface.” SmartCare Mobile is a white-label application that customers can access when they download a participating company’s customer care application. you do not need to go through lengthy dialogues. SmartCare Mobile also has integrated mapping so companies can send geographically targeted advertising messages to interested customers.” Pieraccini says. empowerment is what this is all about. a smartphone platform and application suite for customer care. SmartCare Mobile supports all of the leading global smartphone operating differentiated customer experience. iOS. the company released its first application. “The type of interaction with a customer over a mobile device is so different that it justifies a new division.” he says.” The company’s goal in forming the division was to leverage the smartphone. “One great feature and value proposition to SmartCare Mobile is that it combines a number of customer care applications into one portal. • predefined. and BlackBerry— on a single application platform. billing. text.” That’s one of the strengths of the new SmartCare Mobile app. You can use voice. the type of customer is different. vice president of product management at SpeechCycle. according to Roberto Pieraccini. or touch. Windows Phone 7. according to Pieraccini. “This allows the customer to take control. “It’s mostly 10 | Speech Technology MAY/JUNE 2011 multimodal technology that requires its own systems. notifications. playing games. • a dynamic deployment model. SmartCare Mobile’s goal is “to take calls out of the contact center and optimize the customer experience. “For the smartphone customer. optimizes contact center operations. customizable widgets.

“Smartphones are game-changers for providing world-class customer care.> > N E W S > > T R E N D S > > A N A LY S I S > > N E W S > > T R E N D S > > A N A LY S I S > > N E W S > > T R E N D S > > A N A LY S I S browsing Web sites and are expecting the same rich experience when they communicate with their service providers. If a criminal [hacks into the system] and takes that data. he adds. a veteran software industry executive. “We felt it was necessary to come out with a special guidance document for the call center. including recommendations and best practices. The guidance highlights the key areas that organizations with call center operations must address to process payment cards securely and outlines how best to protect their businesses and customers from the risks of card data compromise.com Speech Technology | 11 .” Gorelov explains.” he says. the guidance expands on a PCI Council FAQ published in 2010.” Kim said in a statement. The council developed the information supplement to help merchants and service providers meet PCI DSS requirements to secure payment data captured within voice recordings. That’s the central message of a new supplemental guidance document the PCI Security Standards Council released in mid-March. • recommendations for merchants when assessing risk and applicable controls of call center operations. But most card issuers have their own penalties for companies that violate PCI standards. nearly two-thirds of them use their mobile phone to contact customer care. “Consumers are carrying around powerful computing devices.” Pieraccini is quick to point out. with tables that map data types to PCI DSS requirements. and that the standards council is not an enforcement body. Included are the following: • an explanation of how PCI DSS applies to cardholder data stored in call recording systems. that creating the mobile division in no way suggests that SpeechCycle is abandoning the traditional IVR market. though. store it for the minimum time possible and make it unreadable with truncation and encryption. European director of the Security Standards Council. said in a statement. Extending these capabilities to the mobile platform is a natural progression for the company. exposing cardholder data to risk. phone-based self-service applications that improve the customer experience while reducing costs. and SpeechCycle has developed a solution to help service providers meet the challenge. “The IVR will still exist. you fall within the scope of the Payment Card Industry’s Data Security Standard (PCI DSS). The division is being housed at SpeechCycle’s New York headquarters and is being led by Albert Kim. “ ” www. too. In fact.” “The launch of the Mobile Division of SpeechCycle is the next step in our evolution. “We have had great success in providing intelligent. A product of industry collaboration and stakeholder feedback. the cost to the merchant is much more substantial than the cost to become PCI-compliant. and we see a real need for it. “Criminals are now targeting call centers because they know there will be cardholder data stored within their systems.” Zor Gorelov. Additional advice is offered on protecting cardholder data in recorded transactions. don’t store it. The attention to call centers reflects changes in the industry and the tactics used by criminals. and • guidance on key considerations faced by call centers when implementing PCI DSS requirements. Its main message is simple. King acknowledges that the laws of individual countries on processing credit card data supersede the PCI Council’s standards.” —Leonard Klie MAY/JUNE 2011 Criminals are now targeting call centers because they know there will be cardholder data stored within their systems. Smartphones have changed how consumers interact with enterprises and service providers. King says: If you don’t need it. Our vision is to enable companies to tap into this powerful medium to build lasting and profitable relationships with their customers.2. “The underlying goal is to protect the cardholder’s data throughout the transaction process. The “Protecting Telephone-Based Payment Card Data Information Supplement” provides actionable recommendations to merchants and service providers processing payment card data over the phone.” King says. with a quick reference flow chart that provides a step-by-step process for determining necessary controls to meet PCI DSS requirements for voice recordings. consumer behavior has undergone a seismic shift. and they expect faster and better customer service.” explains Jeremy King. • specific guidance addressing the storage of sensitive authentication data. “With 300 million smartphones sold worldwide in 2010.” —Leonard Klie Call Centers Receive Guidance on Protecting Credit Card Data If you handle credit card data. And if you do need it.speechtechmag. including suggested methods for rendering data unavailable under query to meet PCI DSS requirement 3. King notes. The PCI standards apply to organizations with call center operations where credit card information processed over the phone can be recorded and stored. CEO of SpeechCycle.

“The report makes it easy to see what’s in the marketplace. and 2011 has gotten off to a strong start.300. it was not as dominant against a handful of politicians.” Other trends highlighted include: • The WFO market has recovered from the recession. and Rutter was third with $21. In a much less-hyped challenge on February 28. Bill Cassidy (R-La. Watson finished on top with $40. and 16.147.> > N E W S > > T R E N D S > > A N A LY S I S > > N E W S > > T R E N D S > > A N A LY S I S > > N E W S > > T R E N D S > > A N A LY S I S TECH INVESTMENTS BUOY WORKFORCE OPTIMIZATION In a new report on the world workforce optimization (WFO) solutions market. “Looking beyond the salesman” and prioritizing price. she says. And while Watson seemed unstoppable against the champions.600. and customer satisfaction rankings should factor into the end user’s ultimate vendor decision. • End users are demanding Payment Card Industry-compliant applications from WFO vendors. Jennings on the last day of his challenge jokingly added a line welcoming his new “computer overlords.). • Desktop analytics is gaining traction and should pick up momentum. The result was Watson.). The WFO suite market.). • Research and development spending on WFO suites is high and is expected to rise as vendors invest in innovation and building suite synergies. DMG Consulting concludes that it’s “a pretty exciting time. • Vendors are offering more featurerich WFO solutions for small to midsized companies. it is even more important now for end users to select vendors that would cater to their individual contact center and IT needs. and this is a pretty exciting time. —Brittany Farb under-the-radar speech news // A team of IBM scientists has worked tirelessly over the past four years to build a supercomputer that could rival a human’s ability to answer questions posed in natural language.). “I’ve been in this market for awhile. Watson. • Vendors are embedding workflow into their applications to make results easier and more actionable for end users seeking to apply the findings. and obscure cultural references that are a mainstay of Jeopardy. At the end of the three-day challenge.” The success of Watson represents a leap forward in natural-language processing— the ability to understand and respond to everyday dialogues. Watson still trounced its human competition. Fluss says. the IBM supercomputer. future direction. Ken Jennings and Brad Rutter. DMG President Donna Fluss says. trivia superhumans Ken Jennings and Brad Rutter. product functionality. slang. and 5. is expected to grow by 7 percent in 2011.” DMG’s seventh annual Quality Management/Liability Recording WFO Product and Market Report forecasts that contact centers will continue to make substantial technology investments in 2011 and 2012. mostly powered by speech analytics. Let’s hope Big Blue gives Watson a celebrity makeover: I’d hate to be the patient who receives bad news about his latest colonoscopy from a metal box that talks through a huge green ball on a plasma screen. • End users are increasingly interested in buying multiple applications from a single suite provider. business analytics. 6 percent in 2012.000. Jennings took second 12 | Speech Technology BY LEONARD KLIE with $24.com MAY/JUNE 2011 . handily beat two storied Jeopardy champions. 15. on February 14.000 manpower-hours to compile. www. Watson had amassed a commanding $77.000. Though imperfect. contact centers.J. IBM envisions a day when medical diagnosis. With a stronger market and increasing options. On the screen where contestants write their “Final Jeopardy” answers. The report took about three months and more than 1. • End users in both large and small contact centers are pushing vendors to deliver hosted WFO and stand-alone applications.speechtechmag. A high amount of innovation and products continue to be introduced. Watson faced four members of Congress: Jared Polis (D-Colo. which competed on Jeopardy against the TV game show’s two longestreigning champions. word play.” she says. and Rush Holt (D-N. complete with the puns. Jim Himes (DConn. compared with the congressional members’ $30.5 percent in 2013. and tech support can be automated by a system like Watson. • An increasing number of contact center infrastructure vendors have entered the market with their WFO suites. It’s a buyer’s market. Based on that success.

“there’s a reason a lot of attention is being paid [to mobile voice security]. and that’s why this year’s gathering in Amsterdam in late March featured security on mobile devices as a central theme. and developers. which is part of the BrightCloud Group. you are in trouble. the event’s organizer.000 top prize. whose BioLock application unlocks mobile phones. About a dozen companies entered the contest. for integrating a solution that incorporates voice biometrics in a way that is easy to use over the phone. Other finalists included Logica.” Miller adds. “It already has your contact list and all your personal information. and it did that and more. according to Miller. verifies people’s identities after receiving locationaware mobile marketing messages. access control.” says Dan Miller. “The future of voice biometrics is now in the hands of adjacent industries. “We do live in a dangerous world.” The challenge originally was touted as a highly visible means to promote awareness of voice biometrics among mobile application developers. which runs on Android-based phones. driven by mobile. and Hold Free Systems.> > N E W S > > T R E N D S > > A N A LY S I S > > N E W S > > T R E N D S > > A N A LY S I S > > N E W S > > T R E N D S > > A N A LY S I S Mobile: The New Frontier for Voice Security A recent Twitter poll conducted by Nuance Communications on consumer preferences for mobile customer service. walked away with the $1. but in the end contact www. More than half (53 percent) of all those polled said that security risks are a main concern when they complete transactions on their mobile devices.com “ ” . user authentication.” —Leonard Klie We’re all exposed to the risk of identity theft every day. senior analyst and founder of Opus Research. BluePlanet Apps. the need for enhanced security [on mobile phones] is critical. wallet. According to Miller. lays the foundation for broader applications and deployment scenarios in the future. the Helex app “captured our imagination and has immediate commercial viability while.” center solutions provider Helex. or other applications. the mobile phone has quickly become “a personal assistant. “It’s still a technology that requires a lot of thoughtful innovation. transactions. Voice biometrics gives them the added level of security needed to ensure that their information is protected every step of the way. and so we’re just now starting to uncover the security flaws in them. PC.” he says. “Consumers are using their mobile devices as their passports to managing their lives. and we’re all exposed to the risk of identity theft every day.” Organizers of the Voice Biometrics Conference agree. which created a mobile banking app. from paying bills to managing and setting up their accounts. According to Miller.” Opus Research “is seeing a broader opportunity.” It’s with that as a backdrop that Opus Research and VoiceVault this year unveiled the Mobile Voice Biometric Challenge to uncover the application or service that makes the best use of voice biometrics on a mobile device for activation. at the same time. With 5 million mobile devices in the world. systems integrators. There’s a real need to have more security on your mobile device and in the apps that run on that device. which entered an app to authenticate remote agents logging on for work at a call center.speechtechmag. and search engine” for most people. and if you lose it. and security found that nearly one-third of respondents believe voice biometrics is the best method for verifying their identities on mobile devices. director of product management at Nuance. 84 percent of consumers contact customer service through their mobile phones. said in a statement. According to the poll. The company’s BioProximity app. “Now more than ever. “There’s an interest among the people that own these devices to do more with them.” Hetal Pandya.

the Department of the Prime Minister. and.” In England. to enhance the traditional approach to voice biometrics by adding a second factor—personal knowledge—to its verification process. for example. or accent of the speaker. Michael Khitrov. calls it “a landmark deployment in the voice biometric industry. and IP phones) regardless of the gender.000) from the U. speaker verification. “When voice biometrics really ‘pops. and transcription of audio from phone taps. has been sponsoring the Biometric Vulnerability Assessment Project since 2007 to develop a principled assessment methodology for the vulnerability of biometric systems to deliberate attack by impostors. Miller points to a “significant amount of investment” in government projects to do speaker identification.” comments Judith Markowitz. mobile. speech-to-speech translation. head of products at VoxGen. Mydex Data Services and Latitude Partners. Technology Strategy Board to research. is being used by more than 250 local and federal law enforcement agencies across the country to match unknown voice samples to a database of known voiceprints to track down criminals. and PerSay [which is now part of Nuance Communications] have been doing for governments for covert operations and surveillance. conveniently. state. securely. under the Research Support for Counter-Terrorism Program. a provider of self-service automation solutions. “There are a lot of areas where the government is investing. In Mexico. “Consumers increasingly want reassurance from organizations that their personal data is secure. VoxGen will develop a statistically rigorous probabilistic reasoning system to combine those two incongruent data points. said in a statement. most importantly.” As a result of that project. “Law enforcement and government are fueling a very quiet market right now. Europe. which the company deployed about a year ago.” he states. according to many industry watchdogs.000 ($481. has been awarded a grant of £300.” Miller concludes.” The lack of noise aside. and Asia in an effort to deploy similar systems. The platform can compare short utterances recorded over various channels (microphone or landline. Though many other voice biometrics vendors have been engaged in similar projects with other governments. “It’s been going on for a long time—you just don’t hear about it. Russian company Speech Technology Center (STC) has been working for eight years with the government to merge forensic technologies and voice biometrics for speaker identification. VoxGen. voice biometrics is expected to gain recognition as a proven identification method with a real-life track record. The resulting voice identification system.” and says the impact of this project “is difficult to overestimate. The main outcomes will be a general methodology for the assessment of biometric systems vulnerability and a specific methodology applicable to a chosen biometric. age. VoxGen will be working with its consortium partners. “For the private and 14 | Speech Technology MAY/JUNE 2011 Government money is flowing into researching voice biometrics. it’s now an economic imperative to have the technological capabilities to identify and verify consumers quickly.” says Dan Miller. Growing demand for voice biometrics in general and voice-based identification in particular signals an emergence of a Across the world. and enhance voice biometrics to support consumer identification and verification technologies. Speech Technology Center. senior analyst and founder of Opus Research.” In Australia. —Leonard Klie www. speech technology companies are stepping up their research and development around voice biometrics solutions at the request of many federal.com “ ” .> > N E W S > > T R E N D S > > A N A LY S I S > > N E W S > > T R E N D S > > A N A LY S I S > > N E W S > > T R E N D S > > A N A LY S I S market spotlight: GOVERNMENT Voice R&D Speaks to Homeland Security public sector alike. STC’s CEO. language. STC is already negotiating with a number of other governments in Latin America. develop. new market segment with a potential to grow to a $500 million industry in a very short time.’ government will be behind it. Markowitz Consultants. “Government money is flowing into researching voice biometrics. language identification. “There’s a lot that companies like Loquendo. the Mexican deployment is the largest known speaker identification project to date. outlining the capability to test and report on the vulnerability of any individual system. and local governments for homeland security and law enforcement.K.” Pierce Buckley. including suggested countermeasures to identified risks.” Government uses of the technology “are much further along then they are letting on and much more advanced than what we’re seeing in the commercial world.speechtechmag. president of J.

European and Brazilian Portuguese. Businesses now can capture. analyze. >> Verint Rolls Out Multimedia Solution Verint Systems has released the latest version of Audiolog for Contact Centers. screen data. call routers and dialers. advanced multimedia recording. including email and text messaging. and playback capabilities. and European Spanish. French. which provides one-click translation for any computer application. cloud-based speech recognition services. tapping on a keyboard. FlexT9 also supports these languages: U. users on all Android phones. and update their Facebook or Twitter statuses. Web chat. email. phrases. and Spanish.S. Mandarin Chinese. two new female voices (Italian and Turkish) are available for TTS. was launched in North America in 2007 with three languages. Arabic. >> Vlingo Aims App at U. a provider of language solutions.> > N E W S > > T R E N D S > > A N A LY S I S > > N E W S > > T R E N D S > > A N A LY S I S > > N E W S > > T R E N D S > > A N A LY S I S SOUND BYTES >> Nuance Crafts App for Tablets Nuance Communications’ multimodal FlexT9 app is now officially optimized for Android tablets. Turkish. European and Canadian French. such as email. as well as an improved acoustic model for U. get directions on Google maps. including select Honeycomb devices.K. operates on a single platform and features new. German. analog. hands-free messaging.speechtechmag. and videos. T9 Trace.com Ford SYNC globally with the European launch of the company’s popular voice-controlled connectivity system. including digital. and Twitter and Facebook posts with their voices and have incoming text and email messages read aloud on Android and BlackBerry phones. has chosen Acapela Group to speech-enable its Babylon 9 translation solution. based on the latest versions of Loquendo ASR (7. and Verizon users with Android phones can download Yap Voicemail from the Android market. The solution. PDF documents. search the Web. Italian. Acapela text-to-speech will offer an audio version of the translation with male and female voices in 18 languages. which combines voice recognition. SYNC with MyFord Touch will be available in U. meaning more than 10 million AT&T. Users can control the device and its apps with voice or by tracing (gliding their fingers from one letter to another on a virtual keyboard). and Taiwanese Mandarin (supported through Mandarin Chinese). and other multichannel systems. Users Vlingo has released a new version of its popular Virtual Assistant localized for U. FlexT9 is currently available for download in the Android market for $4.9) and Loquendo TTS (7. Speech Technology | 15 MAY/JUNE 2011 . The system can recognize 10. Version 5 has been expressly designed for small and mid-sized contact centers with 150 or fewer seats. now offering 19 languages. a technology platform provider enabling voice-activated hands-free communications. and respond to audio interactions. German. email. and other characters on the touch screen). U. and the Motorola Xoom.0. Dell Streak 7. storage. >> Loquendo Updates Speech Suite Loquendo has released Speech Suite. numbers. German.000 voice commands. or writing (using their fingers to draw words. SYNC. to power a new service. and lets consumers switch among any of FlexT9’s input modalities. and quality assurance solution. T9 Write.S. >> Yap Voicemail Fits Android Yap. English. Korean.K. SMS/text. and Vlingo’s proprietary Intent Engine.S. The free application automatically converts voicemails into text. and Microsoft Word files. and XT9. voice-activated. Internet pages. U.1 and higher. Italian. English.20). French. the company’s multimedia recording. and IP phones and PBXs. and U. Japanese. It supports Android OS v2. >> Ford Revs SYNC in Europe Ford is expanding the reach of www.. AVST Team on Voicemail-to-Text Ditech Networks and Applied Voice & Speech Technologies (AVST) entered into a partnership to deliver voicemail-to-text services to enterprises.S. AVST is using Ditech’s PhoneTag platform for automated and hybrid voicemail-to-text transcription services as part of its enterprise unified communications solution. version 7. retrieval. >> Ditech.. and Australian English. The language portfolio also has been expanded with the addition of ASR in Arabic and TTS in Romanian. allows users to speak to their phones to send text and email messages. a provider of automated. >> Babylon Eyes One-Click Translation Babylon.K. backed by Verint’s workforce optimization and recording patent portfolio. The PhoneTag solution is now available in English. developed with Microsoft and Nuance Communications.99. has released a version of its Yap Voicemail for Android. natural language processing. The four-in-one keypad combines Dragon Dictation. and Spanish. SuperTooth HandsFree Assistant users will be able to create and send SMS. including the Samsung GalaxyTab. The Ford Focus will be the first vehicle to launch with SYNC in Europe in 2012. With the language expansion.K. T-Mobile. pictures. Dutch. >> SuperTooth Finds Bluetooth Partner Bluetooth speakerphone maker SuperTooth has chosen Dial2Do. The application. SuperTooth HandsFree Assistant will give users of the SuperTooth HD speakerphone two-way. As part of the white-label arrangement. Russian. archiving. evaluation.

• convenience. Next year. or Z charges. Other motivations include: • scalability. That pattern is expected to continue for at least a few years. • analytics and reporting capabilities. respectively.” DMG President Donna Fluss explains. reflect what both analysts and vendors say is an obvious trend toward outsourcing enterprises’ IVR needs. she predicts. • reliability. we’ll charge two-thirds less. 16 | Speech Technology MAY/JUNE 2011 . during the same period.speechtechmag. one of the primary strategies for some hosted vendors is ‘whatever premises-based vendor X. says Daniel Hong. provided by DMG Consulting. isn’t the only reason that enterprises are turning to hosted IVR. Y. sometimes even less than that. and • IVR optimization services. the consulting firm says. Those projections. followed by a 10 percent drop in 2012. “There are many advantages to using a hosted vendor.’ They can do that because they will reuse a lot of code. “including the fact that they don’t charge as much as their premises-based counterparts.” Indeed. the premises-based inbound speech-enabled interactive voice response (IVR) sector is expected to decline by 5 percent. while revenue from hosted/managed service applications grew by 6 percent and 13 percent. Fluss says. In fact. however. • minimal up-front investment. whereas premises-based vendors typically won’t.Hosted IVRs BRING Peace www. a sector currently estimated at about $500 million a year that is expected to soar past $900 million by the end of 2015. • leading-edge technologies.com T he numbers say it all. the highly competitive nature of the hosted/managed service IVR market and the more than 95 vendors vying for customers put end users in a strong position to negotiate an outstanding deal. Motivations Go Beyond Price Price. But the hosted sector should grow by 8 percent in both 2011 and 2012. lead analyst for customer interaction at Ovum. Sales of premises-based IVR solutions slid by 8 percent between 2008 and 2009.

com MAY/JUNE 2011 Speech Technology | 17 .ofMind Price and flexibility are encouraging corporations to switch from premises-based solutions BY PAUL HYMAN www.speechtechmag.

He cites a scenario in which a company plans to have the IVR for more than six years and “if the enterprise takes millions of calls each day and has the requisite skills set in house to manage a speech application and other functions. optimized for 100 percent availability. a cost analysis. today. “In response to customer demand.” he notes. to really extend the value of the IVR platform.com .” says Dena Skrbina.” Hong says. such as proactive communications. because of platforms that have been inherited from various vendors.” Hong says. If that’s the case. “customers aren’t waiting for the next version.” 18 | Speech Technology MAY/JUNE 2011 www. instead of having different people who need to specialize in each.” Different Vendors. and mobile customer service that differentiates a company from competition by delivering the best possible customer experience while saving millions of dollars every year. if they have the internal expertise.” Fluss agrees that. “Instead of having various management tools. Fluss says. our customers use us as a platform-as-a-service play.” Shirk adds. we have added a third data center. and evolved the hosted reporting system into an analytics offering that provides business-level users with actionable insight to make critical business decisions. added a handset-based Mobile Care application service and an outbound solution.” Enterprises are moving toward IVR consolidation. “The premises-based vendor makes a living when you sign an agreement with them. And so the hosted vendor is much more interested in getting you up and running as quickly as possible because it doesn’t make a living until its customers do. But we find there are a lot of inefficiencies in that model. That point takes on greater importance if the corporation has gone through layoffs and one of its two IVR specialists has left.” Hong says. Ours is very much a Web app model. namely routing and virtualization.COVER STORY “Enterprises just seem to feel more comfortable outsourcing the technology and expertise needed to manage a speech application. So they have actually geared their business model toward servicing their clients quickly and efficiently. That comfort level rises “when the enterprise doesn’t have the internal resources to effectively manage a speech solution. She continues. and when the enterprise is looking to hosting providers for other elements of the contact center. under certain circumstances. When Does On-Premises Get The Nod? While the advantages of a hosted solution are considerable. “especially if you’re not concerned about receiving continuous “Enterprises just seem to feel more comfortable outsourcing the technology and expertise needed to manage a speech application.” she explains. does it always make sense to take that route? Would there ever be a situation in which hosting does not make sense or in which an on-premises solution would be cheaper over the long run? “Absolutely. So. he says. outbound.” says Grant Shirk. Nuance also added ondemand ACD [automatic call distribution] and CTI [computer telephony integration] to the hosted solution suite. “It’s really a cloud-based product. their customer data…all of that stays on their premises just as if we were a Web application. he says. “And once they do that. a premises-based solution is appropriate. Different Benefits Customers who intend to turn to a hosted IVR solution have discovered that various vendors seem to bring different perspectives about the services they provide.” One point of consensus: Before deciding between a hosted speech-enabled IVR system and on-premises solution. Hong points out that managing multiple platforms at the same time isn’t economically efficient. “Hosted vendors only make a living when people access the IVR. director of industry solutions. “Since then. and a needs analysis (see “The Anatomy of a Decision” on page 19). when its call campaigns are seasonal.speechtechmag. In addition. senior director of solutions marketing at Nuance. their inclination is to centralize it all. then by all means premises-based may be economically better than going hosted. and they aren’t paying upgrade fees as part of the license. enterprises are delighted that hosted vendors frequently give them more attention than their internal IT staffs do.” In The Cloud Meanwhile. Instead. their business logic. “They are getting the best of the platform for every call as we roll out improvements. added PCI [Payment Card Industry Data Security Standard] level security. they keep their application. the forward-looking ones who are paying attention to what they can squeeze out of the one IVR platform that will extend across different departments…are the ones who can exploit the advantages of having a premises-based IVR solution. “There are many traditional hosting providers out there who will take your boxes or theirs and put them in a data center and host your application and hook up all the piping. if they have a utilization rate higher than 95 percent. it is vital to perform a detailed return on investment.” Because Tellme’s solution is cloudbased. Microsoft Tellme doesn’t consider its solution a “hosted” one. when it needs to introduce more features or functions but doesn’t have the budget. Nuance Communications acquired BeVocal in 2007 and added features as well as rearchitected the company’s IVR solution. Nuance On Demand provides inbound. For example.

• Infrastructure and operational cost savings. That means continually evaluating their selfservice solutions and iterating to meet customer demands. A well-designed IVR—one that unifies inbound. For many customers. a 1 percent increase in the caller self-service rate equates to millions of dollars in savings. cost-effective employment of rigorous security standards. the on-demand model is beneficial because of the way it prices itself by the minute or uses utility-based pricing. as well as wringing more value out of every interaction. With an on-demand pricing model. and operational performance superiority with unmatched guarantees.speechtechmag. For companies in highly competitive marketplaces. Grant Shirk. In addition. or wireless. such as travel. Today’s business climate and consumers are changing rapidly. a pay-as-you-use pricing model that eliminates capital expenditures spending. On the flip side. it’s imperative that they keep up with customer demand and stay one step ahead of the competition.000 ports of capacity for a full year and pay maintenance on it. and it will pay for that up front. you pay for what you’ve used after you use it. • Continual improvement resulting in higher selfservice rates.” Shirk says. transportation. In most cases.COVER STORY The Anatomy of A Decision N A T U R A L L Y. director of industry solutions at Microsoft Tellme. “The utility pricing model allows them to perfectly right-size their IVR expense against the traffic they’re driving. • Loyalty and higher share of wallet. the ability to more easily handle spiky call volumes. outbound. a variety of well-designed modern technologies. emphasizes these elements in an ROI analysis: • Determine your up-front capital investment versus your ongoing operational costs.” Shirk says. they have infinite burstability. In an on-premises model. When they have an event and go from 5. Successful companies must evolve their businesses and the ways in which they interact with their customers. generating millions of dollars each year with zero agent involvement. they don’t have to carry those additional 4. insurance. The wellknown set of benefits includes refocusing IT staff on the company’s core business. “We have a number of companies on our platform who either have very unpredictable or very seasonally spiky call volume. The up-front cost is basically minor. They don’t have any wasted or excess capacity sitting around. Nuance’s senior director of solutions marketing. weighing a hosted speech-enabled IVR solution against buying an on-premises system requires a thorough financial evaluation. For businesses that are either capital-constrained or have strict policies about when to make capital investments. and mobile service across the customer interaction life cycle—helps companies earn the right to keep and even to up-sell to their customers. Dena Skrbina. Some companies using Nuance On Demand offer their callers the option to buy relevant products at precisely the ideal point in the IVR interaction.000 ports one day to 9. an enterprise is generally trying to do very accurate traffic-volume predictions to figure out exactly how much port usage it would need to buy over time. IVR self-service is evolving into a marketing tool that promotes customer loyalty and spending.com MAY/JUNE 2011 Speech Technology | 19 . that amounts to less than 1 percent of the total solution. financial services. outlines the following three elements as critical to the decision-making process: deployments to improve not only self-service technology but also the effectiveness of every interaction. • Determine your need for scalability. It consists of just what an enterprise would expect to pay for the physical connectivity between the two sites if it is doing any secure data transmission.000 the next. Nuance leverages lessons learned through thousands of www.

Shirk points out. and the scripting can be very intelligent because you know the conversation you are about to have. “So why in the world would you want to wait for the calls to come in when there are so many great uses of outbound IVR—from pharmacy refills to travel company notifications to reminders about your doctor appointments to political calls?” Indeed. is that enterprises must not take their IVR solution for granted.” she explains. “Nuance’s application developers and business consultants partner with our customers to design solutions that maximize ROI. When you look at the evolution of services being placed on top of the core platform and what you can do with them…when you look at the reliability and scalability of the platform plus the other cost advantages. You’re having a single conversation. your annual run rate will generally be more than what your annual maintenance cost would be.” Nuance’s ROI calculator factors in capital expenditures. There’s some very old stuff out there.speechtechmag. she says. SmartListener.” A Compelling Case Ovum’s Hong believes that “all companies should be considering going to the hosted model.” Fluss says. A critical factor often overlooked in traditional analysis is the impact from leveraging business analysts and speech experts that maximize the business benefits of then was able to achieve a 2 percent increase in call automation. you won’t have costs associated with a host. depending on the call volume. In 2011. DMG’s Fluss expects hosted outbound to grow by 13 percent and then by another 13 percent in 2012. very quickly. installing a premises-based solution might be more cost-effective because. after a number of years. or a customer proprietary business metric. for argument’s sake. “It’s a really very good technology that.” The bottom line. “Just remember when you do the analysis that you’re not just figuring in an automated self-service solution.net. you can publish an 800 number and 10 million people can call you for one thing but another million can call you for another 150 things. an outbound call can cost perhaps 10 or 20 cents.” Paul Hyman is a freelance writer based in New York. these types of investments pay for themselves very. it’s usually due to human error. and a customer-derived labor model. Orbitz. hosted should definitely be an option.” Skrbina says. Nuance has a long-standing customer that enhanced its application with one of the host’s technologies. Tellme views outbound IVR service as a proactive way to advance conversations with customers. delivers tens of millions of outbound communications every year. “Hosting certainly doesn’t make economic sense for all companies.” Skrbina admits.” A Nuanced ROI At Nuance.” Skrbina comments. she points out. and cater to the highest customer satisfaction scores.COVER STORY technology enhancements. “Nuance’s ROI projections accurately predicted those savings from deploying our enhancements. But with outbound. “Pretty much anyone who hasn’t taken a look at their IVR environment in the past couple of years really ought to take a fresh look. You need to design for that. when it doesn’t work. There has been tremendous innovation over the past five to 10 years. “On inbound communications. “IVR is a missioncritical solution these days. a single task. Skrbina adds. according to Fluss. enable fast and efficient deployments. shorter call duration. whether it’s higher automation. He can be reached at phyman@optonline. since too many companies believe that IVR is one of those technologies that you can put in and forget about. For example. as well as the expertise to build and continuously maintain the end-to-end solution. 20 | Speech Technology MAY/JUNE 2011 www.” In that case. and Enterprises are delighted that hosted vendors often give them more attention than their internal IT staff does. “Organizations need to realize that if all they can do is get a 1 percent increment in customer usage or adoption of their IVR. “One thing that many people miss regarding outbound applications is that they can be extremely targeted. maintenance fees for hardware and software.” she says. you’re very focused.” he explains. “The enhancement led to a cost savings of more than $2 million per year. “especially not for those that have recently made large investments in speech IVR infrastructure. and not just in the technology but also in the best practices. One Tellme outbound IVR service customer. “If you’re putting in a hosted solution. the cost of the equipment—and there is certainly a large up-front investment—but you also need to figure in your own internal costs of maintaining the solution on an ongoing basis. Some companies will keep their IVR for 10 years or even longer without upgrading it. the company takes a consultative approach and offers companies weighing their decision a detailed ROI and cost analysis.” Outbound Growth To Soar Another exploding trend that is sometimes overlooked is outbound—the fastest-growing sector of the hosted IVR market on a percentage basis. which can be challenging and complex.com .” she says. “People have realized that while a cheap inbound call could cost maybe $5.

Only a converged platform can provide the added benefits to: • Extend (not transfer) calls between systems for costs savings • Convert an outbound voice-enabled call into an inbound speech supported dialog • Confirm speech recognition dialogs with SMS messaging for mobile callers • Eliminate asking for repeated detail during a transfer PLATFORM WITH POWER Voice self service is just one piece of an engagement strategy that companies must employ. Motorists register their license plates and credit cards and can then call the system to indicate where they’re parking. visit http://www. MEMORY AND KNOWLEDGE FOR E-PARKING SOLUTION Parkmobile partnered with VoltDelta to develop a cashless speech recognition and SMS based parking meter system. For example: • History identifies customers and helps personalize interactions based on their order chronicle. and over 30% of callers are in the same space. This patent-pending analysis combined with context sensitivity.com/OnDemand or call 866-436-1169.speechtechmag. helps to ensure that what was recognized actually makes sense as well! Reduced tuning effort and cost is an additional benefit that comes along with customer satisfaction improvements.voltdelta.com .Sponsored Content | May/June 2011 Customer Experience Personalizing Customer Care with Data Driven Dialog: How Leading Service Organizations Deliver Tangible Results with Hosted Speech Technology VoltDelta O OnDEMAND SOLUTIONS R A personalized customer experience is delivered when an automated system remembers key elements from a previous transaction to resolve a query more efficiently. based on what is known about the customer. and call plus agent screen recording address the needs of a more mobile society today. customer segment preferences. The system even triggers SMS alerts when the meter is nearly expired. Contact centers across virtually every industry are realizing customer satisfaction and cost saving benefits from newer technology. such as triggering a special offer if the system knows the caller just moved. More than 95% of callers are in the lot identified. memory and knowledge to automated systems helps anticipate needs and benefits. • Knowledge of customers – Using unique data. VoltDelta’s CrystalWAVE uses multiple audio evaluation techniques simultaneously to filter extraneous utterances and improve overall voice recognition accuracy. Speech Technology Magazine | www. transaction history. It allows companies to employ dynamic IVR menus. virtual agent systems. For more information about VoltDelta OnDemand Solutions. Parkmobile noted that commuters for a particular transit system tended to park in the same parking lots and even the same parking spaces on a daily basis. For example. Engagement encouraged by what is perceived as “automated intelligence” that seems specific to each caller improves satisfaction and contributes to bottom line results. Superior care is also delivered when voice recognition interactivity provides a “heads up” phrase to an agent to avoid repeated questions after a transfer. VoltDelta began by identifying commuters by ANI and then tracked whether they recently parked in a specific lot. A foundation for truly improving customer care with cost savings automation from on-demand systems is changing how customer care managers deploy systems today. VoltDelta’s converged platform ensures that intelligence works across all channels. Capable hosted speech recognition solutions are demonstrating these characteristics today! ACCURATE AND EFFICIENT SPEECH RECOGNITION Voice self-service solutions today are elevating customer care by automating more complex calls and transactions. • Memory of channel. reduces handle times and improves satisfaction. multi-channel customer care ABOUT VOLTDELTA ONDEMAND VoltDelta OnDemand is a global provider of virtual contact center and voice recognition solutions. When access and convenience is the expected norm. preference or location – System “memory” that a caller prefers an SMS message response or “recalling” that a motorist checking a 511 system would like results from the last stretch of highway they evaluated. Meter charges are made on their credit card. MEMORY AND KNOWLEDGE TO CREATE PERSONALIZED EXPERIENCE Applying history. USE HISTORY. presenting information PARKMOBILE USES HISTORY. This “smart” use of data speeds the process. To accommodate these users. Over 2 Billion calls per year are processed within VoltDelta’s on-demand infrastructure. a caller can be flagged as a “New Mover” which triggers dialogs in an automated system for special offers to callers identified as “hyper-buyers” service platforms that integrate core key components such as of voice recognition. or feedback. The application immediately asked if they were parking in the lot and then questioned if it was in the same space they last parked in. avoids calls going to an agent.

speechtechmag.com .Captures Consumer 22 | Speech Technology MAY/JUNE 2011 www.

com MAY/JUNE 2011 Speech Technology | 23 . www. emotional states. and moments in time • BY LEONARD KLIE • Nothing reveals more about a business than the voices of its harshest critics and most ardent supporters—its customers. of customer interactions to unearth the vital intelligence needed to build effective cost containment and customer service strategies. The technology could analyze thousands. And there was a time when basic speech analytics was sufficient to help companies learn what their clients were thinking. even millions.Technologies blend data from different sources.speechtechmag.

” One of the first purely speech analytics vendors to embrace multichannel analytics was Autonomy etalk. the reason for offering such a solution is simple. blog posts and responses. product reviews. blog posts. an enterprise search vendor that blends speech and text data to formulate a single voice of the customer. social media status updates. including audio recordings. Such a blending. and an emphasis on actionable results.speechtechmag. Verint Systems. according to Autonomy. “It’s already happening. companies need to extract actionable information and sentiment from the vastly untapped world of social media. the speech analytics market is as vibrant as ever. and so companies really need to pay attention to what’s in the public domain and make sense of it.” More than 90 percent of customers engage in multichannel behavior when dealing with companies. president of DMG Consulting. either on their own or through partnerships with traditional text analytics providers. external. he says. chat sessions.” she adds. That’s not to say speech analytics has been relegated to a bottom desk drawer somewhere or housed on some remote server that seldom gets used. According to Andrew Joiner. Nexidia.” comments Ed Shepherdson. “Now we can go outside a company’s firewalls to get information that is contained within the public domain. email and documents. realtime capabilities. a provider of enterprise search solutions that recently launched the Customer Experience Optimization (CXO) solution to blend information sources. On the contrary. Outside the Walls With these new capabilities to pull voice and text together. president of Vivisimo. and there’s spoken communications. from $95 million in 2009 to $180 million by 2014. called multichannel analytics or analytics convergence. for example. review sites.” Fluss says. Innovations.SPEECH ANALYTICS However. wiki entries. news articles. and forum comments. it is gaining ground. the days of basic speech analytics are long gone.” says Donna Fluss. Speech analytics takes care of phone conversations and other audio. the same customer who called in a complaint to the contact center might have searched for products on the company’s site and commented about the company on Twitter. And if you want to fully understand what the customer wants. Slightly more than a year ago. “There’s written communications. and CallMiner—are also providing multichannel analytics as part of their solution sets. survey responses. but I also send a fax or write an email. you have to look at both. like the Web and social media. point-of-sales information. transaction records. and public data sources. which include text analytics integration. Utopy. with research firm Ovum predicting it would nearly double. partnered with Clarabridge to add Impact 360 Text 24 | Speech Technology MAY/JUNE 2011 www. given that the foundation of speech analytics is the transcription of calls from spoken to written words and then the indexing of those words in relation to other words and throughout the entire recording. emotion detection. chat threads. “A lot of business today is being done in unstructured data.” The blending of the two is a fairly easy process. chief executive officer of Autonomy’s Promote Multichannel Technology Business Unit. Web site visits. Web forums. Naturally. videos. the company released Explore. This form of analytics. When it comes to the voice of the customer. Thanks to changes in how vendors position the technology. while text analytics handles all of the written forms of communications. “all content comes through a single search box. reducing the time needed to find information. managing director of customer interac- tion solutions at Coveo. Speaking in Text Once marketed as stand-alone solutions. Twitter posts. CRM records.com . which lets businesses connect to and understand almost any type of customer interaction. Explore has more than 400 available connectors to internal. speech analytics more often than not is being packaged with text analytics to provide a 360-degree view across all of the ways in which customers communicate with and about companies. email. Verint. For instance. are driving adoption of speech-based solutions in the contact center and beyond.” explains Kevin Calderwood. “You need to be able to control all the mentions of your brand. and social media. including text messages. “speech is just one of the channels involved. Other traditional speech analytics vendors—including Nice Systems. makes it easier for companies to determine whether there is a correlation between what’s in the public domain and what they’re getting from customer surveys and from the call center.” he says. “is absolutely where the industry is going. “I call on the phone. “I expect the market to move more and more in the direction where one company offers both speech and text analytics in one solution set.

“But it’s probably in an early stage. “the ability to put speech into context is critical. Michael Maoz. One school relied on acoustic qualities—such as tone. dominant terms. but without context.. volume. if it’s hosted or on premises. and sighs. Utopy. which might indicate a gap in the agent’s knowledge about a particular subject. Coveo’s approach is to look at how the words are being spoken and used. calls the move toward more real-time capabilities the technology’s natural progression. In no way will the technology ever be superior to a human’s abilities in that regard. to determine the emotional state of a caller.” Fluss observes. hesitations. Advancements have taken the technology from basic acoustic-only or linguistic-only solutions to more advanced solutions that depend upon both models. and intensity— while the other looked at linguistic elements.” Jeff Schleuter. and after the emotional response occurred). The technology needed to perform this kind of analysis is still fairly new and in some circles might still be called bleedingedge.” According to Lomanto. pitch. and other factors—it could take a few hours to a few days to index the audio. vendors have been busy enhancing their speech analytics applications with sentiment analysis technologies that can identify more than just when and where a particular word or phrase appeared in the dialogue.” he said in a recent report. for example.” he says. interest in multichannel analytics has been strongest in the hospitality. texts. “We look at whether the company is being spoken about in a negative or positive way by looking at all the words around the one that you’re looking for. as in so many others. but the real value is in the trending data.” A major step in that evolutionary process has been the melding of the two schools of thought when it comes to speech technologies for detecting emotions. Depending on the volume of data— where it sits. as more and more vendors experiment with real-time analytics. chats.” Schleuter says. inflection. etc.” Shepherdson points out. Talk-over analysis also can identify periods of silence during calls. That requires applications that can identify when customers have expressed an emotional response and can then connect those responses to a particular sentiment. “Everyone wants information. there’s also email. email messages. “There’s a lot of attention and investment in emotion detection. and insurance industries.” he says. Lomanto agrees. and significant relationships among distant ideas using multitiered relevancy modeling. retail. research vice president for customer strategies at Gartner.” says Diego Lomanto. contact centers process thousands of calls. according to Schleuter. notes. “We’re seeing a push for text from customers. One of the main reasons for that has been the very nature of speech analytics. the pauses. that looks at the words being used and when those words and phrases are employed with other words and phrases. laughter. But that is changing. vice president of marketing and business development at Nexidia. but it has evolved considerably recently. and online chat sessions with customers daily. “We’ve definitely graduated our speech analytics to offer more customer interactions in one interface. Nexidia uses more of a lexicon approach. but now with our new offering. but he’s not ready to call it a trend. That capability identifies where in a call the customer and agent are talking simultaneously—a common indicator of customer dissatisfaction and frustration. It’s definitely something that the industry is pursuing.” Therefore. which typically involved recorded conversations that are taken apart after the event.SPEECH ANALYTICS Analytics to its Customer Interaction Analytics portfolio. Autonomy’s Intelligent Data Operating Layer (IDOL) derives meaning by determining patterns of information. “It’s all part of an evolving set of solutions. and technology sectors. “because so much is written about them every day. during. pitch. it means nothing. “Real-time speech analytics is a technology that has been waiting for a market breakthrough. but also to correlate those variations with the linguistic content (including what the caller and agent said before. such as the words used. solutions marketing manager for speech analytics at Verint. Another way in which speech analytics technologies are being applied to emotion detection involves talk-over analysis.speechtechmag. looking at spoken interactions as they occur. speaking rate. And while those interactions hold a wealth of valuable business insight. With Feeling In those industries. “There are call recordings. social media.” but the growing desire for such solutions has also come from the financial services. healthcare. www. where more is being hyped than is in adoption. stops.com MAY/JUNE 2011 Speech Technology | 25 . has seen interest building for text analytics and speech analytics together. “You want to be able to search for a word. I Want It Now! Speech analytics is still far from perfect when it comes to detecting and responding to human emotion. uses this twodimensional approach to emotion detection to not only pick up variations in tone. and surveys.

of particular interest is Nexidia’s Classifier features. This has resulted in an even better experience for all our customers. a data warehousing company. rather than waiting a few days or weeks to get the data.” Schleuter says Page One Ventures is “doing something unique with how they use our technology to improve their operations. Page One incorporated Nexidia’s Enterprise Speech Intelligence suite into Page One’s operations in September. 83 agents manually screened the recorded greetings.SPEECH ANALYTICS “When fully integrated with customercentric solutions. Coveo can supply this data in real time via dashboard widgets that open on users’ computer screens. principal support analyst at IBM-Netezza. “People need to react to changes in their businesses very quickly. to have dynamic views of our customer relationships. “You can set how often you want it to go in and pull data.” Fluss says.” he continues. a supplier of management and technology solutions for social networking providers.” Vendors are taking a look at real-time capabilities “because their customers are asking them for this.” Verint’s Lomanto adds. we can provide them with immediate access to the information needed to make informed business decisions and respond to the customer in seconds. wrote in an email. It can be very close to real time. all the way up to our CEO. “We’re working on our first implementations now. we can update it very quickly.” “Speech analytics is definitely starting to find its rightful home. and they need to react in nearreal time. Processes and measurement metrics will need to evolve for this to happen.” Jeff Prete. “For real time. but the value is clear: Uncovering customer intent and gaining insights during the actual interaction enable organizations to deliver exactly what customers need in real time. Page One Ventures has lowered agent costs by more than 60 percent while screening every greeting to ensure the best possible user experience. which in turn led to a 67 percent increase in the number of customer bugs fixed. “Once we have an initial index. said in a statement. and those greetings are screened to ensure that calls were routed to the appropriate agent. Nexidia customers also can take advantage of some real-time capabilities contained within the company’s Enterprise Speech Intelligence suite. was one of the first companies to use those widgets. and other speaker characteristics in real time. Prior to working with Nexidia. it will enable contact centers to realize their strategic business potential.” but adds that implementations like this one are the future. IBM-Netezza reduced the time needed to identify known problems by 67 percent and cut the number of duplicate bugs submitted to its development engineering department by 50 percent. Instead of spending hours manually compiling reports from data in multiple systems. You don’t have to wait for it to be turned over to the knowledge management solution or for the voice to be converted to text. improving satisfaction and loyalty. they could cover only approximately 5 percent of the total volume.” Coveo’s Shepherdson says. “Coveo dashboards and the analytics they provide are allowing our managers and executives. It is accomplished by triggering and presenting to the agent context-sensitive recommendations and information according to pre-set business rules. which can detect language. “Nexidia’s technology helped us in several areas. “where they are not just culling data but also making it more actionable. and ultimately driving revenue growth. “Now using Nexidia speech analytics. “To be able to analyze calls as they are happening and help agents improve performance and serve customers better while they are still on the phone” will be important moving forward. Now Page One Ventures automatically reviews all caller greetings in real time and can flag relevant calls for further revirew. it comes down to a single one: “You have to mine the data more quickly so you can do something with it. It’s about maximizing the results. Real-Time Guidance provides next-best action recommendations to the agent in real time during a phone call or chat session with a customer. Fluss says speech analytics in general is in the early stages of its second generation. “The information is constantly being re-indexed as soon as something happens. Callers who dial into Page One Ventures are required to record a short greeting before entering each session.” Guidance on the Fly Nice Systems was a pioneer in realtime analytics when in March 2010 it released Real-Time Guidance as part of its Nice SmartCenter suite of speech analytics offerings. This is the key to securing the customer relationship.” he adds. “This is particularly important when one of those accounts becomes hot with an issue that requires immediate attention. gender. general manager of Page One Ventures. As a result.” Today’s solutions pull together the data at much higher speeds and are far more dynamic. the technology is there. It requires much deeper integration into the telephony environment and more tieins to the CRM system. The reasons are manifold.speechtechmag.” Jim Coleman.com . while the cycle times for processing the data are getting shorter and shorter. 26 | Speech Technology MAY/JUNE 2011 www. Within 30 days of the implementation.” One such implementation is at Page One Ventures. including our quality assurance process. He continued. but even working 24/7. but in the end.” Shepherdson points out. Schleuter says. we are able to dramatically increase the volume of calls reviewed and conduct quality monitoring of our contact center with a much more rigorous set of data. IBM-Netezza.

com/europe Learn how speech technologies are impacting on the way we do business today and how to build them into a whole range of activities and environments. London. including: > Customer care > Call centres > Security > Gaming and entertainment > In-car applications > Public transport > Medicine and healthcare > Warehousing and supply chain > Telecommunications > Mobile applications > Public sector applications > Financial services Speakers from: > Google > Barclays Bank > Deutsche Telekom > Ovum > Nuance > Loquendo > Openstream > Voxeo Learn about: > Business strategies > Speech biometrics > Multichannel applications > Multilingual applications > Multimodal applications > Assistive technologies > Analytics and Measurement > Voice User Interaction design > Speech application development tools and languages > Case studies and more 40% Produced by S DISCOUNTES GU FOR COLLEA . can technology provide the answer? Alex Waibel looks at language technology solutions that address language and cultural barriers. UK 2011 Keynotes Hear Google UK’s Engineering Director. particularly for Android and Chrome. Over the past couple of years Google has been making big investments in cloud-based speech recognition services and applications. In a world where everyone could connect if we shared a common language and common understanding. Hear Professor Alex Waibel on providing solutions that bridge the language divide. Dave Burke discusses and demos some of Google’s future plans in this area.Register online at www. on Google’s plans for cloud-based speech recognition for mobile and web. and demos some brand new cross-lingual communication systems. www.com/europe Moving forward with speech Conference & Expo 25 & 26 May 2011 SpeechTEK University 24 May 2011 Copthorne Tara Hotel.speechtek.speechtek. Dave Burke.

28 | Speech Technology MAY/JUNE 2011 www.speechtechmag.com .

our o to pr rgan otect izatio its in n do vestm ents? Off Course s a consultant working with a Chinese language learning company called Qooco. Qooco found itself trying to establish relations. he felt confident when Qooco opted for Big Blue’s speech recognition technology in 2007. However. While the decision to go with Big Blue did not cost any jobs.speechtechmag. in January 2009. problems ensued. and attract attention from strangers. influence product design. Karl (Bill) Scholz understood the conventional wisdom that “nobody ever gets fired for buying IBM. our attention was split between enhancing the system A www. “For a while. his feelings turned to discomfort as IBM sold off its business to Nuance Communications Inc.BY PAUL KORZENIOWSKI Don't Let M&As Steer You Tech vend ors g et ac What quire can y d.” So.com MAY/JUNE 2011 Speech Technology | 29 .

When a new market. president of NewSpeech Solutions and AVIOS (the Applied Voice Input/Output Society). For the customer. suppliers grow because there are so many new accounts. While that statement sounds logical. However. vice president of voice technology at ATX Group.com . the penetration rate increases. It has become difficult—some might say impossible—for one vendor to develop all the necessary components. As a result. what can customers expect from such transactions.” notes Thomas Schalk. it is common for vendors to speak with customers and reseller partners about the potential change well before it is announced. Rather than servicing the customer. depending on the size and complexity of the deal. starting with why the vendor bought the company. so keeping and enhancing the product becomes a straightforward decision.SURVIVING A TECHNOLOGY ACQUISITION and trying to figure out how best to interact with Nuance. acquisitions have become a common way for vendors to expand their businesses.speechtechmag. As those changes unfold. We Love Techies Once a deal has been signed. That move depends on several factors. it may not play out in reality. the other OEMs would not want to have part of their sales buttressing their competitor’s bottom line. But customers often feel the ripple effects from such changes. cashing in on the fruits of their labor. the ground is level and start-ups sprout. more congested boardrooms. In other instances. Since purchases are expected to continue for the foreseeable future. Last year brought more than a dozen purchases (see “2010: A Year of Change. Like other high-technology sectors. “Acquisitions help suppliers fill any product line voids. the market matures. Another factor in the merger mania is that speech is becoming a common user interface. They race to create buzz and build market share. however. speech is volatile. before the sale is announced. Eventually.” notes William Meisel. in the case of start-ups. In fact. Suppliers have focused on filling out their product lines. the decision process is not as clear-cut. and. They revise their resumes. buds. They have to alter how they interact with their primary vendor. grow significant market share. In an industry rapidly evolving. They do not want customers and partners learning about the merger by reading the latest headlines. acquisitions should remain common in the speech industry. and what can they do to protect their interests? Part of the Circle of Life Because consolidation is a typical phase in an emerging technology’s life cycle. However. Initially. however. 30 | Speech Technology MAY/JUNE 2011 www. the choice to discontinue a product is also simple. Many speech technology purchasing decision makers find themselves in similar positions. a telematics provider. If the technology is an item in an emerging market and there is no overlap. Because large companies tend to be bureaucratic. the most important decision the purchaser makes is determining what will happen with the acquired technology. vendors often make acquisitions to help round out their product lines. research director at Gartner Inc. and they end only after the buyer has fully integrated the acquired company into its operation.” For customers. the time needed to answer a customer’s call can lengthen. Sometimes. “Usually. an acquisition’s impact is felt first on the personnel front and later with the underlying technology. employees worry about how the potential purchase may affect them. a variety of parties is involved in the process. and find a way to have their voices heard in larger. then the product usually will continue to be supported as the combined company moves forward. senior manager of solutions marketing at Nuance.” says Jeff Foley. This process can drag on for months or even more than a year. internal politics overrule “Acquisitions help suppliers fill any product line voids. the acquirer has little incentive to invest time and money to help other vendors find uses for the technology. a start-up needs to find a defendable niche. call friends about job openings.” page 31). The distractions begin weeks. For instance. the new company keeps the technical personnel. rumors circulate. or find a buyer in order to survive. If product overlap exists. “Speech solutions no longer stand by themselves. the client may have to interact with new sales and support personnel. In that case. they are being integrated into many other products. problems arise because not all parties keep their mouths shut. has made more than 30 acquisitions—including nine in 2007— since the turn of the millennium.” notes Steve Cramoysan. As the chatter grows. Company principals try to keep news of a possible buyout as quiet as possible. Moreover. and opportunities become scarcer. Sometimes. In fact. they often cannot move fast enough to adapt. try to safeguard equipment that is often sent to the scrap heap. driving down costs. such as VoIP [Voice-over Internet Protocol] systems and unified communications applications. president of speech consulting firm TMA Associates. the buyer may start to cut staff. or even months. a start-up developed a novel technology and then OEM’d it to various large corporations that are direct competitors. and check newswires for the latest information. such as voice biometrics. and Nuance Communications Inc.” says Scholz. “the best option wins.

Standards. To differentiate their wares. Mapping Out the Future While that scenario may not sit well with customers. there is a good chance that the number of purchases will rise.2 million for Norway-based Global IP Solutions.” says Elizabeth Herrell. Nuance purchased MacSpeech.” says Dan Miller. —P . a supplier of investment banking services to the technology industry. the more difficult it becomes to swap one system out for another. And in January. Theoretically. Usually.com MAY/JUNE 2011 Speech Technology | 31 . Rogers Wireless. so the different solutions will have a similar code base. which had developed voice search applications.to software-based systems. Back in 2006.K. Two months earlier. and the user was often forced to scrap its existing system and start anew whenever an acquisition occurred. www. The buying frenzy is expected to continue this year. there were 3.’s growth. including Bell Canada and Vodafone Turkey. As the new company is assimilated. Apple acquired Siri. the market is shifting from hardware. Traditionally. For instance. “While it may not be simple. in fact. according to Regent Partners International Ltd. West Corp. Acquisitions have been a key component of Nuance Communications Inc. and speech companies have been a lively segment of this space with double-digit purchases. whose focus had been on carrier voice recognition systems. The economic downturn blunted. a migration path to a new system. if necessary. First. “Most vendors continue to support the current product at least until a new release combining both companies comes to market. making it more difficult to port it to the new environment. the larger supplier will lay out a road map that outlines future product developments and. 2. such decisions are not uncommon. so the migration challenge is being mitigated. In the end. was also active. The more special functions they add. the acquirer’s product will win the battle for acceptance. they wait until the current system needs to be refreshed or when it no longer meets performance criteria. president of market research firm Communication Initiatives. so customers need to view acquisitions as likely when they purchase their next speech solutions.672 purchases took place. senior analyst and founder of Opus Research. West had purchased HollyConnects. In July 2010.295 acquisitions in the hightechnology industry. in some cases. the standard typically provides only baseline functionality. Nuance ponied up $102 million for SpinVox. In May. speech suppliers built their products atop proprietary hardware. Nuance acquired PerSay. standards are not a panacea. they will be given time and maybe some help from their new master in moving off of their current solution. In 2010. but did not stymie. While they can help clients. “It is in the best interest of the purchaser to keep the current customers happy and migrate them to its equipment. for a couple of reasons.’s Macintosh line. the vendor acquired TuVox.” ATX Group’s Schalk notes. a provider of speech and interactive voice response applications. The growing importance of speech in business and consumer applications was illustrated by moves by the high-tech industry’s most prominent players.” Schalk says.. are designed to help businesses jump from one system to another. and its customers. they may have a solution that is a 2010: A YEAR OF CHANGE A cquisitions have become an important way for vendors to fill voids in their product lines and grow their businesses. “Often. Combining the different solutions presents the vendor. a VoIP/unified communications provider. the merger movement. How to Move from Here to There Increasingly. NewSpeech Solutions’ Scholz worked with a VoiceXML-compliant system but found that a vendor had extended its systems. making two purchases. with challenges. which made voice recognition solutions for Apple Inc. the vendors work from common design points.speechtechmag. companies can move from one system to another more easily now than they could in the past. Google forked over $68. In February of that year.technical merit. such as VoiceXML. manufacturers often layer features. a speech recognition supplier whose clients include Bell Mobility. a voice biometrics supplier with about 200 customers. and the company made three of them last year. “After an acquisition. In April. the customer faces the decision of buying into the new product road map or moving in a different direction. In December.” Communication Initiatives’ Herrell says. and Vodafone Spain. companies will opt for a new system. If the economy rebounds.

If they bank on a start-up. “Established companies.SURVIVING A TECHNOLOGY ACQUISITION few years old and struggling to keep pace with new demands. They may not be willing to divulge their financial information. so the customer has to determine how much effort it wants to extend. health care. Nortel. chances are good that the organization will change—dramatically— over time.com. instead. either by being acquired or as a natural part of its growth. Google. When an industry Goliath takes over the product. Often the smaller company has developed a niche item that complements what the larger vendor already offers.speechtechmag. That’s because there are costs associated with doing business with a variety of vendors.” If a customer isn’t safe buying Avaya. He can be reached at paulkorzen@aol. Such wording is now fairly standard. Customization may become easier.com . the client should describe its desires and concerns with the relationship. Reviewing those items requires time and money. since the acquirer will probably want to minimize disruptions at the beginning of the assimilation process. integrated solution. In addition. Relying on a Start-up Is Risky Corporations should be aware of the risk they are taking. such as a corporate restructuring or the possibility of being purchased. vendors make decisions that they view as their best “Large Fortune 500 companies hand integration and migration worries over to their partners. In addition. Customers should speak with the management team first to make certain the vendor knows who they are. they need to be aware of the potential challenges and do what they can to protect themselves. customer input on a potential purchase is limited. so enhancements are in line with its needs. it must go through a complex legal review before agreeing to the vendor’s terms and conditions. In addition. IBM. meeting the new owners is recommended. or financial services. the change can streamline a customer’s business processes.” option. then there is risk in every product purchase. the customer may be offered a “sweetheart deal” to stay. such as telcos.” Scholz says. Apple. If the vendor is acquired. “Large Fortune 500 companies hand integration and migration worries over to their partners. like Avaya. Opus Research’s Miller notes. Paul Korzeniowski is a freelance writer who specializes in technology issues. or Siemens. Finally. Given the recent downturn. If a client has unique vertical industry requirements—say. The reality is that no customer is secure. Taking that step helps to forge a relationship the way the client wants it. In the event of an acquisition. So. financial services’ regulations about call recording—that client often has to tweak the speech system itself to support such functionality. Looking on the Bright Side While there are some downsides. the client needs proactively to monitor the relationship and keep its ear to the ground for information that could influence its interactions. The buyer may have a large reseller network. resellers often target vertical markets. the reseller may isolate the customer from any product migration issues. The company can roll any acquired technology into its existing line and create a more complete. Also. 32 | Speech Technology MAY/JUNE 2011 www. Consequently. and Siemens. So dealing with fewer vendors reduces the volume of work in such areas as procurement administration. the client should be mindful of opportunities to influence the development process. In the end. Since acquisitions are so common. customers also can benefit from acquisitions. Finding financial data is simple with large public companies but gets more difficult with smaller private enterprises. Trying to discern their health can mean doing a lot of legwork ferreting out the data. someone else may take on that work. businesses should take steps to protect themselves. Improving Business Processes In some cases.” Nuance’s Foley notes. While the fundamental elements are similar. Each time a company adds a supplier.” Opus Research’s Miller points out. If a vendor is struggling to pay the bills. Another way a business can protect itself is to include language in the original contract giving it the right to renegotiate the terms if an acquisition takes place. the customer gains a better understanding of the vendor’s product strategy. Businesses cannot ignore the market realities. the focus and wording of those agreements vary from one vendor to the next. have gone through dramatic restructurings. “The customer should inquire about and track the vendor’s financial health. putting itself up for sale is one way to solve the problem. While the industry giants focus on horizontal sales. and Nuance are billiondollar enterprises that acquired smaller speech suppliers in the past year and dramatically increased the reach of their new products. Nortel. “the bigger vendors can allocate more resources and spend more money promoting the product than the smaller company. many entities have established a corporate objective to reduce the number of vendors with which they work. taking a step back to put a new solution in enables the customer to move a few steps forward and add needed functionality.

speechtechmag.We’ve got a lot to tweet about.com MAY/JUNE 2011 Speech Technology | 33 . Join the conversation! @SpeechTech SCAN HERE Smartphone Interactive Users www.

Florida. Texas. the Washington-based Scripps Howard News Service.>> SUCCESS STORIES IVR Is Good News for Scripps Media company leverages Pronexus system to standardize subscription services for 14 newspapers | BY LEONARD KLIE he E. Scripps Co. Indiana. Florida. Ohio. and puzzles and games to more than 400 editorial offices around the country. and Washington. Kentucky. Maryland. Kansas.” columns from nationally syndicated writers like Miss Manners and Cokie Roberts. and the United Media syndication service that delivers comic strips like “Dilbert” and “Marmaduke. Tennessee.000 weekday and 790. and Arizona and one independent TV station. Newspapers account for about 56 percent of total revenue at Scripps.speechtechmag.W. 34 | Speech Technology MAY/JUNE 2011 www.000 Sunday subscribers in cities throughout California. Michigan. operates 14 newspapers. The central Florida–based media company must manage a readership base of 630.or NBC-affiliated local television stations in Oklahoma.com . Missouri. South Carolina. which also owns and operates nine ABC.

Lindoo says.” he says.” Jordan Sommerville. and tor built the subscription system based credit card payments will be handled on VBVoice.speechtechmag.” Lindoo says. Ed Lindoo. Scripps has begun standardizing all of its newspaper subscription services on a single interactive voice response (IVR) system built using the VBVoice development toolkit from Pronexus. of IT infrastructure at Scripps since The current IVR usage rate falls be2001. THE SCRIPPS CO. other Scripps newspapers and 60 customer service representatives have started to adopt Lindoo’s IVR. Pronexus. and we get terrific support from www. I time zones and vary its greetings based could have it done in a week or two. which was made by matter of about three or four weeks. we Chatterbox. ning on that one system.” The new systems are being hosted at Verizon collocation facilities located in New York and San Francisco. systems at each location expected to change much with a single centralized syswith the new system because tem. delivery confirmations. VBVoice is free for developers. Lindoo says. Lindoo has modified multiple sites. VBVoice can help deliver superior customer service while keeping the costs down.” Lindoo says. “If I was The new system will cover three just doing a single-site IVR upgrade. and Lindoo expects similar savings for the rest of the group.” multithreading. of the nature of the calls being handled. the increased customer retenScripps’ VBVoice-powered The Scripps IVR tion and cost savings that it IVR accepts touch-tone inis expected has already yielded. noting the project involves Over the years.” he says. MAY/JUNE 2011 Speech Technology | 35 .000 calls a month.000 calls a is fast. ” App At a Glance SINCE IT BEGAN INSTALLING AN IVR BUILT WITH PRONEXUS’ VBVOICE TOOLKIT. we’ve a 70 percent increase. It’s an economical price point. and the average call to field about lications are also replacing handling time is now two to 10. and credit card processing. Once they have been completed. which isn’t a day. which is based in Ottawa. but on the local time. just kept building on it. “It’s been very easy to use and very reliable. We will not have to worry about take about 18 months and is due for comthat anymore. both the central IVR and the central cir“The difference will come when we culation system will be cloud-based. and savings on the project reach $1.” he says. the Scripps IVR is VBVoice is its flexibility. Canada. “because some people just don’t tracted by the system’s affordwant to talk to a machine no ability and effectiveness and matter what.000 calls separate circulation database three minutes. databases.” The company employs between 50 Recently. back-office circuthe system with easy-to-use drag-andlation systems. Making changes expected to field about 10. atas well. “In a fiercely competitive market like the newspaper industry. HAS SEEN: IVR usage climb to between 57 percent and 59 percent. The pubputs. has been operating at behad a fully functioning IVR working tween 32 percent and 35 percent. once the system is ready for deployment. senior director through the automated systems. for user interface-based VBVoice at its Treaall 14 of the Scripps newspapers. drop controls several times in response “It’s a pretty big undertaking because to changes in other back-office systems we’re going to have 14 newspapers runand in Scripps’ IT infrastructure. All desure Coast newspaper group in central livery complaints. or about 200. while further development of it.>> SUCCESS STORIES >> SUCCESS STORIES >> SUCCESS STORIES With subscribers scattered throughout the country. “Within a the old system. “One of the things that I like about Once completed. vacation stops and Florida in 2002.5 million a year. giving the Scripps IVR the capacity to handle up to 250 calls simultaneously. have a call center that’s slammed with The entire standardization project will calls. Scripps first installed the graphical day.com “ It’s a pretty big undertaking because we’re going to have 14 newspapers running on that one system.5 million a year. “The new IVR will save us somewhere around $1. up from between 32 percent and 35 percent with its previous system. After a systems integrarestarts. VBVoice has been hugely successful for the Scripps publications that use it. customers would pay run-time licenses based on the number of lines they have. said in a statement. “From there. territory sales manager at Pronexus. That’s for us. took over the maintenance and tween 57 percent and 59 percent. taking advantage of what we’re doing is pretty complicated. pletion sometime in early 2013. and each channel can handle as many as 125 calls at a time.” Lindoo says.

“I pay an annual maintenance and licensing fee. Pinal County’s customer service administrator. but it only recently improved service to residents by implementing a communications-asa-service (CaaS) solution from Interactive Intelligence. 42 government agencies routinely interact with citizens through the Citizen Contact Center. 36 | Speech Technology .000 CALLS A MONTH.>> SUCCESS STORIES >> SUCCESS STORIES >> SUCCESS STORIES Rural Arizona Connects With Residents inal County in rural central Arizona is home to 325.m. and we came up with the numbers to justify it. My costs have become much more fixed.000 and 15. which gave us the server capacity for an on-premises solution. Internet reliability. From its county seat in Florence.” says Jerry Keely.m. but cost was the biggest factor.000 TO 15. who notes that the contact center’s seven agents field between MAY/JUNE 2011 Pinal County speeds service and cuts costs by switching from a hosted to an on-premises contact center suite | BY LEONARD KLIE 10. he was paying on a per-minute/ www.000 residents who are as diverse as the area’s landscape.speechtechmag. call and screen recording. which involved changing from a hosted service to on-premises. The system enabled the Pinal County government to add features such as skills-based call routing. which is open Monday through Friday from 8 a. and supervisory monitoring. and call quality issues that plagued the county’s previous system. The implementation.” Keely says.” Bringing the technology on-premises removed the distance. a speech-enabled interactive voice response system. “We started watching our call volumes. “Cost is not something I have to concern myself with anymore.com THIS BUILDING IN FLORENCE HOUSES PINAL COUNTY’S GOVERNMENT.000 calls a month. “And we had just completed a data center for the county. The center has been operating since June 2007. to 5 p. began a little more than a year ago. BUT AGENTS WORKING OUT OF THEIR HOMES CONNECT RESIDENTS TO THE RIGHT EMPLOYEES BY FIELDING 10.” Keely explains that with the hosted model.

.......... and seen its first-call resolution rates jump the number of calls answered within from 70 percent to 87 percent........... according to Keely...13 www..speechtechmag....... and other data SINCE SWITCHING FROM A HOSTED CONTACT from its own servers to the new onCENTER SUITE TO AN ON-PREMISES premises servers in about three weeks. he says the solution was perfect for the county at the time.” an on-premises version was an easy process..........infotoday.. we can continue to imhealth emergency...............W.. I received my bill and I did not know what those costs would be. Now. “Hosted althe average wait time is 114 seconds.. Interactive Intelligence moved all of Pinal County’s App At a Glance IVR scripting....... tion as needed.....000 and $15. Instead............... and lowed us to expand without being tied to the abandon rate is 15 percent............ Cover 3 www.. “I was expecting to have ismonthly costs shrink to $300 a month sues.worldlyvoices...... 325.. SOLUTION.... turned the system over to its Public Health “As a result of Interactive’s innovative Department for use in the event of a capabilities. 33 www.............com 615-321-8802 ............avios.. 15 Pinal County (Arizona).....speechtechmag... “I was paying $10............... In fact................ but prove on our efforts through faster rethey can open it up and turn it on in a mosponse times.. Cover 2 www... either............com 503-645-3598 Speech Tech Best Practices Series ... residents to contact us..... ent server....larson-tech.............. adjust routing in real time system to better but the difference now is based on fluctuations in call serve Pinal County’s that agents log onto a differvolume....... and improved Switching from a hosted solution to agent training.............. additional ways for ment’s notice if needed....... We could set up at “The Interactive Intelligence CaaS other locations if needed and Contact Center solution has offer our agents to work enabled us to automatically The decision was from home........ as well as a hosted to a premises-based home-based agent approach........>> SUCCESS STORIES >> SUCCESS STORIES >> SUCCESS STORIES in this issue “ Cost is not something I have to concern myself with anymore....” from between $10.” Keely says..” Keely says.000. The average www.... Scripps Co.....com/perodicals 800-300-9868 Larson Tech ....... so we went with hosting so we could grow the solucall handling time is now 213 seconds. Pinal County has first-call resolution rates jump from 70 percent to 87 percent. 33 www............. 33 www. 34 Ford Motor Co.. “At the end of the month...... “It’s not live now.......com SpeechTEK Europe 2011 ............... we didn’t know how much call volume we would get.......com ” ■ SPEECH USER COMPANIES References are to the first page of the story or section in which the company appears. my basic circuit costs are about $300 a month....” he says.... 36 ■ ADVERTISERS AVIOS...... 27 www.... routing................. a particular location.. ......... 33 speechtechnology @ speechtech SpeechTEK 2011 .....SpeechTEK....... Since the transition.... per-call basis......speechusability..... 5.........” he adds...” Keely explains.speechtechmag..... Cover 4 www....... E....... and the 20 seconds rise by 50 percent..” connect residents to the right made to switch from The county still has a employees faster...000 residents... but it was flawless.............. THE PINAL COUNTY GOVERNMENT IN ARIZONA HAS SEEN: “The transition was amazingly simple............” But Keely has not soured on the hosted model....com/subscribe Speech Technology Twitter .... it county is a priority for us.. “Since this was all brand new to the county.....................com/webinars Speech Technology Subscriptions ............. Offerings .................” he says................com/europe Speech Usability .000 per month in telecom costs with hosted.....000 to $15.........org Information Today Inc. “Providing superior servPinal County did not abanice to the citizens of our don its hosted solution.SpeechTEK.com 512-897-0350 Worldly Voices. number of calls answered within 20 seconds has climbed 50 percent.... 7 www.... 6.

fact that it is designed to use the H. post your ideas to the the public mailing list.com potential to M ultimodal applications have incredibleand applicaenrich the interaction between users .efficiency. As discussed in the MMI archito use in an application. nested. Sample components include nents and the interaction manager. and defines error messages tion. nents. report its http://www. A complex component combines two or experts to supply components to integrate into multimodal more distinct functions that are bundled for convenience or systems. because this involves a high-level architecture.HTTP for transmitting media. tion.org/TR/mmi-mcbp/. You can find the MMI architecture spec at That is. An example would be an avatar that needs to comification (MMI architecture) is one such standard. and video display. components can be simple. may include an internal interaction opening the door for smaller players. To make it easier to design interoperable compo. guidelines ease design of processing modalities (modality compoFinally. together to create innovative applications.example includes information such as the fact that the ity Components. and resume. scripting languages. Ph. for multiple third-party components can be problematic. They involve the internal create independent versions of each technology they want architecture of components. The third type of components. manager.D. The handwriting component architecture. and biometrics. pause.the face-recognition example defines error messages such as nents. dahl@conversational-technologies. Simply mastering separate APIs tecture spec. handwriting recognition. stop. audio formats for speech Conversational Technologies and chair of the World Wide Web Consortium’s Multimodal Interaction Working Group. the MMI architecture face recognition. But multimodal capabilities. such as “codec not supported” if it is asked to start displaydesigners need more specifics to create pluggable compo. communication protocols such as HTTP.org/TR/mmi-arch/ and the best practices status. The eight guidelines should be included in the specifica. like speech recogni.org. and error codes also are important to 38 | Speech Technology MAY/JUNE 2011 www.Those three examples also include specific XML examples tion of a component so that third parties can use it in MMI of life cycle events used with each example modality.architecture for each component. or nested. as well as the other life cycle spec at http://www. Similarly. Think of the guidelines as a The MMI working group is interested in hearing your “spec for a spec” for modality components. the W3C Multimodal Interaction Working Group “known users not found” if the database of previously recently published a companion document to the MMI enrolled faces is not found. the best practices document has interoperable nents).w3.264 codec. someone who’s building an application and considering using another vendor’s compotions. ity component must implement all of the life cycle events. or complex—is appropriate. The definidard set of life cycle events. events defined in the MMI architecture. She can be reached at recognition. tions describe APIs based on the MMI By standardizing interfaces between modality compo.com.ing a video file that uses an unsupported codec. Simple components are internally black boxes. The Multimodal Architecture and Interfaces spec. but in Standards for integrating technologies from different some cases it may be useful to create components with vendors would make it easier for third-party technology internal structure. are themselves The three design suggestions are also helpful for designers highly complex. But. at The first. coordinated by a central interaction three examples of definitions of modality comcomponents. is principal at speech and language consulting firm Supported natural languages.DEBORAH DAHL STANDARDS Making Modalities Play Nicely Together MMI architecture opens door to collaborative multimodal apps define. handwriting recognition is based on InkML input and that three design suggestions. complex.w3. employs cific functionality. it must accept commands to start. any events specific to the component must be defined. lets vendors offer pluggable components that can fit into The video display component explains such details as the other companies’ multimodal applications and provide spe. feedback. Deborah Dahl.www-multimodal@w3.nent can know whether it would be suitable.speechtechmag. it uses EMMA for representing the recognition results. nested. and few companies have the resources to of interoperable components. The best practices document gives Eight best At its essence. and most important. architecture-based systems. With that information. called “Best Practices for Creating MMI Modal.” Included are eight important guidelines.. guideline is that a modal. such as speech or handwriting recogni. In addition. gesture recognition. the MMI architecture suggestions for deciding which architecture— practices defines a set of cooperating components for simple. The municate closely with a TTS to synchronize lip movements architecture enables various vendors’ technologies to work to the synthesized speech. and three examples of components. manager and communicating through a stanponents that follow the guidelines.

or SGDs. transportation. Los Angeles has been ordered ers. Manual Communication Board: With the puricans with Disabilities Act and other federal and state laws. nothing can substitute for personal reached at (888) 999-9161 or contactus@comptalk. and home kits. Individthem.ROBIN SPRINGER VOICE VALUE Emergency Preparedness Is a Job for All Court ruling in Los Angeles disabilities case teaches a broader lesson creating their own plans. “They may not even know nication barriers are being released regularly for smart there is a table. such nearly 1 million Angelenos with disabilities. as the availability and variety of mandated on a governmental level.com).” phones and tablets. Replacement and/or Loaner can take more time…But as the scale of the emergency For its part. “Specifics need to be in place as to what can be are forced to leave home without medication. SGDs. instead of traditional assistive technology (AT). abilities should get involved and take part in community “During a small-scale emergency. manufacturemergency preparedness was being decided in California. mobile apps increase. and emergency shelters.” Kailes says. a lawsuit about unless a source of electricity is available. Alternative Power Supply: Powering devices during our personal emergency preparedness plans and those of extended outages is a major concern for all people.comptalk. gency preparedness plans and.providers of the equipment would be reimbursed. traditional AC. which include coverage for ment or loaner SGDs.’s RehabilFrom word prediction to visual voicemail.com MAY/JUNE 2011 the horrific suffering caused by natural disasters W ithJapan and New Zealand saturating news outlets in Speech Technology | 39 . as well as car. government Robin Springer is an attorney and the president of Computer Talk Inc. and where users can obtain the replacement SGDs and how Speech-generating devices. redundancy is key. chase of each SGD. Each one response scenarios (such as in ambulances. USB. at least in Los Angeles. preparedness.” 3. agencies now can alert people of impending emergencies. manufacturers should provide notification.speechtechmag. first responders can take emergency response training. a consulting firm specializing in the design and implementation of speech recognition and other hands-free technology services. and shel. more users may rely on these products It’s good advice all of us can use. because so many people have mobile devices. that set forth criteria and the availability of replaceto create those specific plans. Manufacturers can be part of the solution by addressing the following issues: and social media sites. and individuals should be proactive in www. Thus. including gosingle SGD to communicate across multiple emergency bags.” president of Augmentative Communication Inc. are used by non-ver.Enhancement (AAC-RERC).” though it consists surance companies should have polioptions are available.” disability policy should have policies. While traditional SGDs may always be Being prepared for emergencies and disasters is now critical technology for some. Specifics should be included on how people with communication difficulties. violated the Amer2. ERs. Experts agree that it may not be feasible to build a uals should compile multiple emergency kits. “But as the scale of not being invited to the table. that the city failed to provide for the disabled in its emer. first responders of an emergency or disaster. She can be Nonetheless. our cities and states.should include alternative power supplies and hard copies ters) because the devices are highly customized to the indi. Blackstone.com. emergency. productivity itation Engineering Research Center on Communication software to text-to-speech. as such. holding that Los Angeles failed to laminated manual communication board for emergenensure accessible emergency evacuation. the emergency escalates.” mostly of general information on how cies in place to provide prescription people with disabilities should be medication to disaster victims who treated. cies. bal people. not as many options are available. the suit claims as battery power. in conjunction with insurance carriconsultant June Kailes says. new apps that overcome commu. car lighter adapters. ers should emphasize the importance of alternative power Brought against the city of Los Angeles on behalf of the sources and provide options for charging the devices. instructions on how to create your making the disabled “disproportionown manual board. “[People with disabilities] are more time and be creative. Or. SGD makers done by whom and in what situations. many of us have begun to examine 1.and even solar-powered bags. bedside. though non-native speakers might also employ As for personal preparedness. Almost simultaneously with the AT. manufacturers should include a The court agreed. the city argued that it Policy: Just as pharmacies and inescalates.” says Sarah W. Moreover. People of all vidual. at a minimum. (www. and mobile devices that require power are useless earthquake in Christchurch in February. But SGDs may be one of several tools. not as many does have a “plan.of important documents and phone numbers. for inclusion in “During a small-scale ately vulnerable to harm in the event emergency kits.

writer who understood how science is done. a science fiction regime the regulators would impose. biometric companies would scramble to explain might struggle. and wait with dread to hear what nonsensical security formulated by Poul Anderson. As for the hypothetical vast opportunities. Eventually. honest people’s T 40 | Speech Technology MAY/JUNE 2011 www. public would improve people’s lives (that is. would have a party on their hands. lives). the costs of compliance would burn money that everyone who visits Web sites regularly would click through might otherwise be spent creating real value. the more complicated it becomes.g. which would is nothing but trouble. And the behaviorists are thinking that industry.. with enforcement by formance of your speech technology and user interface. I tried for hours I know what all you technical people are thinking: What to think of them but to no avail. all the data would offer insights into the perproposed legislation to protect privacy. IVR systems without thinking through the consequences. As usual. to avoid high complirect whatever background marketing information you’ve col. abide by the FTC’s Fair operations are so innocuous that we should be exempt. Even straightforward IVR systems might stinkyjournalism. or. data collection practices into a rigid straitjacket. As for our some sleepless nights. The more I look at it. Must Information Practice Rules. face higher compliance costs. I’d likely be The Obama administration in March proposed legislation liable if company A breaks the rules). At first I considered that a nightmare! A hundred million Web unique visitors to your some companies would rather shut down their Web presence site. I do not prefer the color red”). and problems and trends. opening 100 million records to the scrutiny of the general He can be reached at speech@pobox. bags of cash to beg for an exemption? Data mining compa• right of review to verify that your data are correct and nies.ance costs.and provide telephony services instead. Maybe it would do both. each one wanting to exercise the right to review and cor. primary source materials (see http://www.D. Ph. complicated. open up vast opportunities. if I use company A’s API in my product. the proposal higher costs through a regime of audits. for the mill as you grind away at your user interface and backThe paper trail starts at the Commerce Department. metrics but also to let consumers review that Even News reports about the legislation have data. that would result in ways we can’t begin to predict. the disclosure agreements. New Web-based how engineers work: “I never saw a problem. Given that the law would require that data straightforward echoed some spokesman’s talking points be presented in an easily understood format. including the ones that sift through corporate data for complete. so I dug struggle. Entrepreneurs where might sell an extra biometric system to provide secure who want to analyze and monetize statistics would be in for access.php?id=1263). if ever. such as Web APIs. any API-centric platform would ther. is president of Disaggregate Consulting and author of The Pebble and the Avalanche: How Taking Things Apart Creates Revolutions. perhaps the legislation would translate into permanently lower start-up activity. Any service using biometrics The European Union advocates similar reguwould be required not only to secure the biolations with similarly unclear consequences. This in turn would require: you disclose all the uses of this data? How can you accom• precollection disclosure of what data would be plish that during a phone call? Do you have a lobbyist with collected from you and for what purpose. that when looked at the right way wasn’t still because of privacy rules and the sheer cost of compliance more complicated. the FTC would The point of the legislation. Security folks would marvel at the cloddish notion that Moshe Yudkowsky.. Every user of your IVR system would provide grist around for the full story and the government press releases. would become harder to create. give references to features to the man on the street. however services. • security of all your personal data. At first glance.” (e. which end technology. That alone would that would affect the Internet and our professional lives in squash experimental APIs. able to wave our arms and claim that our data collection in this case Internet-based services. but I doubt that would happen. the Federal Trade Commission. We won’t be The government would be able to mandate that someone.com.MOSHE YUDKOWSKY FORWARD THINKING If This Passes. is to force enforce rules through its panoply of coercive measures. if I read it correctly.speechtechmag. Only Criminals Could Do Speech Technology Federal privacy bill would drive up compliance costs and inhibit market oday I’m reminded of what I’ll call Anderson’s Law.com . as well as Some consequences are easy to foresee. Someone somelected (“No. Or neiAs for speech technology. rendering the operation pointless.org/editordetail. Journalists rarely. a few people would sell extra IVR systems.

Speech Technology helps organizations design. Get your free subscription today! www. and news that drive the industry forward. and deploy the latest speech technologies as effective and efficient methods of interacting with customers and employees. Covering the people. .speechtechmag. implement.Speech Technology magazine has become the voice of the rapidly growing speech technology industry.com/subscribe Smartphone interactive users SCAN HERE. events.

. NJ 08055 • Phone: (800) 300-9868 . Some magazines are available digitally and are free to all qualified subscribers. Inc. Smartphone interactive users SCAN HERE. family of magazines invites you to subscribe to any or all of these magazines for FREE*! Visit us online today at www. Inc.S. *Print editions available free to qualified U. subscribers.com/periodicals for links to these and other Information Today. magazines. Information Today.infotoday. • 143 Old Marlton Pike • Medford. Inc..The Information Today.