Nice, well-thought-out assessment. While a lot of focus has been on user growth — and the stock price reflects this — some analysts believe that comparisons to Facebook are flawed and that MAU is the not the best metric by which measure the company’s health. So, before we propose solutions to drive user growth, we first have to accept the premise that user growth is a problem or at least is one that should be prioritized. I’m not sure it is.
But, assuming we accept that premise, the consensus seems to be that the obstacle to attracting new users is around “Discovery.” First, I’d suggest taking a step back and thinking about what makes Twitter appealing. Why were you attracted to the service and why do you continue to use and enjoy the service?
If the value of Twitter is content based on on topics — or what I’d call your “interests” — and events, then you’ll likely gain more value from it by following people you don’t in real life vs. people you know. Why? Because not all of your real-life relationships are based around your interests — heck, most probably aren’t. And therein lies one of the more appealing aspects about the service to me: *I’M* the curator on Twitter as opposed to Facebook where people in my real-life social graph are the curators. You and other Twitter fans and power-users likely inherently understood this and, therefore, embraced the work required to curate your own feed.
If you agree with the above, then the challenge is that while Twitter’s value is based on interests and events, the service is structured around following people. So how do you know who to follow? Hence, the Discovery problem you expounded on so well. But Discovery and curation go hand in hand. Attempts to make Discovery easier inherently mean trying to automate curation, which is bound to be flawed. How will Twitter know that the “best” person for me to follow could be some random person in Idaho who is passionate about a similar interest? Of course, once Twitter knows what that similar interest is, it could easily figure that out. But how will it know? Back to your point about the “cold start” problem.
I do think there are ways to make Discovery easier, but I’m skeptical of approaches that rely on editorializing content since, again, part of Twitter’s value is that I’m the curator. Interesting things could be done with hashtags and objective analytics, but it still requires the user providing some small, meaningful piece of data in order to get meaningful content in return. And part of me wonders if that’s just the price of admission and inherently why Facebook-like MAU metrics are neither realistic nor necessary for the service to be valuable to users, advertisers, and investors.
