The Cognitive Startup

by Sarah Dillard

“Listen to your customers, but don’t do what they say.” Eric Ries

I’ve been on Quora fewer than ten times.  My impressions of it are thus snapshots, and I have not grown gradually with it.  Each time I’ve visited Quora, I’ve been impressed by the quality of answers to questions about the Silicon Valley ecosystem.

But on the non-Silicon-Valley topics I care about, the questions still aren’t good, let alone the answers.  One only needs to browse through “microfinance” or “emerging markets” for ample evidence.  Meanwhile, each time I go back, the feature set is larger and more confusing—I have followers?  An inbox?  Notifications?  I thought I was just looking for an answer to a question, maybe one I didn’t even know I had.    

I’m not privy to Quora’s iterative process, but it feels like a company that is listening to its current user base rather than thinking about what it will feel like to join as a new user.  For Quora to “cross the chasm,” it will have to be intuitively usable by new users—and it feels like it is getting further from this the more it iterates on its current product without adding new content verticals. 

There are two popular camps in decision-making: the Malcolm Gladwells who Blink and the Michael Maboussins who Think Twice.  In How We Decide, Jonah Lehrer unites the two, explaining the neuroscience behind the decisions where it makes sense to go with your gut and those where it makes sense to step back and think carefully. 

The customer usage data collected through rapid prototyping and iteration is essentially building a limbic system—a gut—that gives feedback on how customers are responding to the product.  But a gut is not sufficient.  Without help from the frontal cortex, the part of our brain we recruit to help us address new situations, the learning machine may suggest regression to the mean or an increasingly elaborate set of features that meet the needs of beta users, as it feels like Quora is doing.  If the limbic system helps you listen to your customers, the frontal cortex helps you decide what to do about what they say.

Startups have historically been long on vision but prone to errors in assessing adoption and usage patterns.  The rapid decline in the cost of building prototypes and collecting data on usage patterns is beginning to create the opposite possibility—that startups  might have little sense of who they want to be and how they add value but reams of data about their users.  A “cognitive startup” gets both pieces right—it has a limbic system that gives it rapid feedback and a frontal cortex that can interpret that feedback.

Dropbox is an outstanding example of a cognitive startup.  As founder Drew Houston comments:

We have features that our power users have been requesting from the beginning that we still haven’t added.  One of the most humbling things we’ve done is bring in people we found on Craiglist to try our product.  They struggled to use it.  We learned then that we need to keep it simple.

And simple it is.  My mother, a physician who into the 2000s didn’t know of the existence of search engines, is already a Dropbox fan.  In its current form, Quora has no such chance. 

But all is not lost.  As Jonah Lerner ends How We Decide, “The most astonishing thing about the human brain [is]: it can always improve itself.  Tomorrow, we can make better decisions.”  Tomorrow, your startup can become a cognitive startup.  And so can Quora.