The LAB: Monetizing Twitter with Attribute Dependency (February 2009)

Venture capitalists could increase the value of their investments by applying a corporate innovation method to those investments.  Take Twitter for example.  It just received its third round of funding – $35 million.  Yet it has no revenue, no business model…just the promise of such.  It is the perfect time to innovate.

I decided to take the challenge to create new concepts for the Twitter platform that have the potential to earn money.  Others are chasing this, too, including the Twitter management team.  It reminds me of the early days of Amazon when many (including me) wondered if the company would turn a profit.  The difference between Twitter and Amazon is an important one.  Amazon started with a business model in mind.  From there, it had to achieve economies of scale.  Twitter started with none.  Economies of scale do not matter until it can define a viable business model.

Let’s see how innovation can help.

I used the Attribute Dependency template of Systematic Inventive Thinking, a method of innovation that works like no other I have found.  Attribute Dependency (or AD for short) differs from the other templates in that it uses attributes (variables) of the situation rather than components.  It is a powerful tool and more challenging than the others in some respects.  It yields amazing results.  You start with an attribute list, then construct a 2 x 2 matrix of these, pairing each against the others.  Each cell represents a potential dependency (or potential break in an existing dependency) that forms a Virtual Product.  Using Function Follows Form, we work backwards and envision a potential benefit or problem that this hypothetical solution solves.  Innovation!

Let’s start with my attribute list:

• number of people followed
• number of people following you
• number of tweets received per day
• number of tweets sent per day
• length of message sent
• type of message sent
• demographics of sender
• relevance of message received
• emotion of sender while tweeting
• sensation of sender while tweeting
• current activities of sender
• who the sender is with
• location of sender
• why the sender is at that location (context)

You can download the actual matrix I used, including my handwritten notes, here.  From this, I created several interesting dependencies, and then grouped them together to create new service offerings for the Twitter platform.  Imagine a new premium subscription level of service called Twitter Smart Sense.  It would do the following:

1.  Twitter Tagging:  Twitter scans every tweet and tags it based on the linguistic syntax within it.  Tagging is based on tense used in the tweet (past, present, future), demographics of sender (gender, status, age, etc), current location, status with the receiver, and state of mind (happy, sad, bored, excited, etc).  The receiver can then opt in or out (filter) messages based on these tags.  For example, “I want to receive tweets only from happy-minded females in Pittsburgh who are family or friend status thinking about what they are going to do in the next 24 hours.”

2.  Auto Complete:  Twitter fills in details of your tweets automatically based on your location, time of day (as a proxy of where you are – work, home, airport, etc).  This is like the auto-complete function in certain Windows applications.  It remembers the nature of your previous tweets and “nudges” you on what to tweet about.  It does the thinking for you.

3.  Sender Tagging:  Imagine a Twitter appliance (iPhone app?) that allows you to pre-tag your tweets with your current mood, geo-location, demographic profile, etc.  You have the option of sending this out to a specific set of receivers based on how the tweet is pre-tagged.  This helps match the tweets to those you are following who are most likely to be interested in it.

I invented these three concepts using only a small sample of cells in the matrix. I invite readers to use the same matrix (or create their own) and offer their own ideas.  The key is to select a cell in the matrix randomly, then imagine the new dependency.  Think of it this way: “As Variable 1 changes, Variable 2 now changes, too.  How would that work?  Why would it be useful?  Who would think that was useful?”  Using Attribute Dependency takes some practice, and this exercise is a good way to start.

Many think of Twitter as “life streaming.”  I see it more as “relevance streaming.”  That is where the money is.  The option to send and receive tweets that are highly targeted and meaningful raises the importance of Twitter in our daily lives.  This is certainly the direction the Twitter community is headed based on the steady stream of new Twitter applications.

To follow me on Twitter, click here. (I am in a good mood…I promise!).