w w w .w s j 2 . c o m
cal. Collaboration requires sharing repre-sentations for processes that once wereproprietary, and this is not an easy step totake. But if companies are under pressureto be better, faster, and cheaper, they willhave to do only what they do best. Whatever a company doesn’t do well has tobe done by someone else. Hence, thegrowth in process outsourcing and thesourcing of external services. If commerceis to be truly collaborative, the underlying business processes must collaborate too,both within and across firms. This must beachieved at the business level and from thetop down, leveraging existing systems inthe enterprise. That is, collaboration muststart from the business purpose, not thetechnical constraints.
What business needs is not a one-timefix for individual processes but a connect-ed systems environment that can flex andrecombine as required by changes in themarket. Most companies now want morecontrol over their own processes, moreinteraction between their processes andthose of their partners, and some controlover and monitoring of processes per-formed on their behalf by partners. Firmsare also seeking to expose discrete busi-ness competencies as processes they cansell to others or through channel partners.To do all this, firms need to understand theprocesses that underpin their core busi-ness competencies. In short, they need aBPMS capability, not a new suite of enter-prise applications.The situation is similar to the periodbefore the invention of the relational data-base management system. Business dataused to be embedded in applications. Asthe volume of data grew and the connec-tions between data sets in different appli-cations became important to the business,it became obvious that data should bemanaged outside of the application archi-tecture. For example, managers wanted toanalyze the data for business performanceindicators. To achieve this, the new meth-odology for data management was built ona formal model called the relational datamodel. Today, this is commonplace. By allowing a company to manage its dataapart from the applications that use it, thedatabase management system (DBMS)supports a variety of data models and datamanagement tasks and tools. The IT indus-try as we perceive it today is largely found-ed on the DBMS. Today’s enterprise appli-cations are primarily concerned with read-ing, writing, and manipulating data tables– in other words, clerical tasks. This has aprofound implication – such applicationsare stovepipes. Business logic, data model,time, and connectivity all exist within theindividual application. Creating and man-aging end-to-end processes has, up tonow, depended upon complex middlewaresolutions. Not only is this expensive, it’soverly complex.New process management systems of-fer a potentially simpler, more cost effec-tive, and manageable alternative. Oneimmediate benefit will be the ability toalign processes more directly with organi-zational objectives. Business processes lit-erally define the firm and represent thesource of all competitive advantages andmarket differentiation. Business processesare complex, long lived, unique, numer-ous, and constantly evolving. As the drivetowards automation, process outsourcing,and collaboration continues apace,processes have the potential to overwhelmthe firm. Process management systems aretools for managing that complexity.In this climate, standard processesdelivered in the form of standard applica-tions that are also available to competitorsare less and less attractive. Businesses want to shape their processes themselves,perform continuous and incrementalprocess improvement without impedi-ment from technology, and simultaneous-ly exploit low-cost best-of-breed applica-tion components. Powerful new processservers will support this approach, provid-ing a hybrid environment combining thebest of component application engineer-ing and the best of process engineering.The era of stovepipe applications willeventually give way to the era of processmanufacturing.
A Process Langhuage
The first step is to make processesexplicit by abstracting them from applica-tion software. This is hardly new. Decadesago, operating systems were created by abstracting memory management, fileaccess, and graphical user interface from
In the world of alliances and ‘virtualcompanies,’ where many or even mostof a firm’s staff may be working on jointprojects with other businesses, whatmakes an appealing partner? There is aquick and easy checklist concerned withproducts, prices, market access, finan-cial muscle and track record. But anincreasingly dominant question is, whatis it actually like to work with these peo-ple? How do they do business? Is theirway of doing business likely to provecomfortable and compatible for us? Acompany's processes are becoming akey criterion in the beauty parade forglobal partners. Is it possible to doubtthat the quest for a universal processdescriptor language will be successful? Itis increasingly hard to doubt that thequest will take place, and will commandsubstantial interest and resource duringthe next decade.
CSC Research Services
It's no surprise then that BPM isquickly emerging as the moniker for thenext killer app in enterprise software.Few areas of software will receive moreattention in the coming months andyears than BPM. Yet the greatest chal-lenges to the BPM market are the veryforces making it so attractive.
Enterprises should begin to takeadvantage of explicitly defined process-es. By 2005, at least 90 percent of largeenterprises will have BPM in their enter-prise nervous system (0.9 probability).Enterprises that continue to hard-codeall flow control, or insist on manualprocess steps and do not incorporateBPM's benefits, will lose out to competi-tors that adopt BPM.
For the Fortune 2000 companies, thequest to implement the best businessprocess management (BPM) solution isbecoming highly desirable - akin toacquiring the "holy grail" in any givenindustry. BPM promises to streamlineinternal and external business process-es, eliminate redundancies, and increaseautomation.
What The Analysts Are Saying