Twitter's Revenue Model

Because being cool doesn't pay the bills forever.
Jonathan Ludwig Jon@JonathanLudwig.com

What's working for Twitter?
•  Simplicity o  Simple web interface and simple features o  Character limit creates accessability •  Growth o  More users means more visbility o  More visibility encourages more users, including corporations •  Early Adopters o  Branded adopters are authorities -- CNN, Scoble, TechCrunch, NY Times, JetBlue, etc. o  User base is active elsewhere online, high visability

Twitter needs to be stable
•  User Experience o  Changing how basic functions work or adding complexity might threaten the loyal users and could find them moving to other services o  Stability and availability of the service have been problems in the past, these cannot be sacrificed (now that they are much better!) •  Nature of Content o  Twitter should maintain it's relatively spam-free status-pushing unsolicited content would be bad o  The value for most people isn't in commercial uses for Twitter, the value is in social uses--commerce is just moving to where the conversation is

The Opportunity: Corporate Tools
•  How each company should use Twitter is different, it's not one size fits all (e.g. buying keywords) o  Twitter shouldn't try to solve this problem. o  Twitter should empower companies to solve this problem themselves. •  Tracking and measuring effectiveness is currently an issue. o  As a company on Twitter, one knows how many followers one has and which links get clicked on. o  More granular information is needed. •  More flexibility for interfacing with Twitter is vital. o  Getting data into and out of Twitter should be easier o  Using the API is too complex a solution for simple problems.

Example Feature: Tracking/Statistics
Currently, there is no way to check how successful a tweet is in being recognized. A corporate account could look at, and download data on, how successful an individual tweet was at being seen and "passed on"

Example Feature: User Search
Corporate users will have the ability to target groups of users based on engagement. Also, tracking growth of communities would be possible. Benchmarking community size and activity relative to competitors is another potential use.

Other Potential Features
1. Products that allow more complex interaction with Twitter. For example, the ability to download statistics as Excel files or RSS feeds, the ability to Tweet information by FTP'ing files, via email, or through plugins for common programs. 2. More integrated tracking for links, including aggregation across all tweeted links, breakdown by common web metadata (geography, platform, etc.), and where the link was viewed (API, web, mobile, Twitter search, etc.). 3. Enhanced API to allow corporate users to access this data in proprietary software. For example, Salesforce could integrate statistics from Twitter into its suite of reporting and analysis tools.

On The Subject of Pricing...
Here is the proposed pricing (also discussed in the FAQ). Initial Pricing Flat Fees: $500 / month, initial three months are free Usage Fees:   $500 for the ability to exceed API limits   $500 any month where there are more than 1,000 updates   (including DM's) Long-Run Pricing Flat Fees: $1,000 / month, possible tiered pricing scheme Usage Fees: Tied to Twitter's estimation of resource utilization (e.g.   Google App Engine or Amazon Web Services)

Revenue Projections, Part 1
Assumptions •  There are 100 corporations or other branded entities on Twitter, willing to pay for corporate accounts, starting right now •  The number of paid accounts will grow by 30% annually (baseline growth) •  80% of paid accounts will not incur usage fees •  The remaining 10% will incur usage fees of $1000 / year, with that number growing by 25% / year (less than the community growth/baseline growth) Twitter's actual growth: 750% per year Seems like my assumptions could be very conservative...

Revenue Projections, Part 2
Default Assumptions Year 1: $470k Year 5: $3.6MM Year 10: $14.3MM Aggressive (50%) Baseline Growth Year 1: $470k Year 5: $6.3MM Year 10: $51.9MM

FAQ 1: Competition
Can't someone else implement this, perhaps better than Twitter? I highly doubt it, and there are two reasons. 1. Some of the features I suggest would require either new data to be collected and stored by Twitter or Twitter to expose data that isn't already available. While this poses an obvious technical challenge, it's also impossible to duplicate. 2. Twitter has limited the API, and for good reason. However, this make it impossible to do a truly complete analysis when trying to target users, statistically gauge reach and effectiveness, or access other data.

FAQ 2: Usage
Wouldn't a free account be sufficient for corporations? Probably not. If a corporation wants to manually interact with Twitter (certainly this is required for some uses of Twitter) then a free account will continue to do that. However, this requires resources--the tools mentioned earlier would help optimize the use of these resources. Further, even the most manual tasks, when targeted, would be more effective. As corporate use of Twitter moves away from just sitting a person in front of a computer, more features than free accounts offer will become much more vital.

FAQ 3: Pricing
How did you come up with your pricing scheme? Why is it correct? My pricing scheme was driven by two fundamental principles: 1.  Pricing should promote growth and experimentation 2.  Incremental costs from increased use should be covered The price point for a corporate account is low (initially free for a short time) to encourage corporations to sign up and spend time finding the best use of Twitter. Initially, since corporate users' usage might differ from a standard user (heavy use of DM's, higher percentage of @replies, etc.) a flat fee for potential strain on the infrastructure of Twitter might be needed. In the long run, as Twitter's value is proven and corporate use matures, it'll be easier to anticipate usage costs and a higher pricepoint for corporate accounts can be more easily justified.

FAQ 4: Other Ideas
What about pushing content? selling ads? charging per tweet?
Pushing content to users degrades the user experience. Also, one endearing feature of Twitter is that it's difficult to spam users--this model flies in the face of that. Selling ads that aren't targeted in any way is of dubious value--past experience has shown these to be ineffective and irrelevant to most users. Further, placing ads on the web interface cuts against the simplicity and ease of use, and pushing ads through the API is impractical. Charging based on tweet/impression/follower/<whatever> is exactly the wrong idea--each scheme presupposes a corporate model for using Twitter. Charging involuntarily will also discourage experimentation and requires paying without any more functionality than a personal account.

FAQ 5: More Features
What if companies need tools you haven't thought of? As companies begin to get creative about how they use Twitter there will obviously be more features they want or need. For Twitter, developing features will be another way to expand their profit opportunities--the anticipated return on investment for each requested feature will guide development. Maintining a regular dialog with their paying corporate customers will also provide Twitter with a wealth of new ideas and potential innovations that can help advance the service.