Tldr: Twitter’s product strategy is a function of their funding strategy as much as any other driver.
Congratulations Twitter. You’re clearly relevant and vital enough to generate an incredible amount of dialogue around what you could (or should) be when you grow up. Media company or tech platform? Walled garden or open API? Free to users or paid service? And it’s not just hand-wringing; smart folks like Dalton Caldwell are betting that there’s appetite and need for something different.
The various analyses provide varied explanations for the twist and turns in what Twitter has become, and where they’ve signaled they are heading. One ‘camp’ lost and another won. Or something like that. For me, the ignored elephant in the room, and the biggest driver behind the current strategy, is not who happens to be CEO or what’s philosophically more ‘pure,’ but rather their fundraising history and the expectations which come along with aggressively raising money.
In summer of 2011, Twitter reportedly raised funding at an $8 billion valuation, after several other rounds during the previous years. This heady sum didn’t make them ‘over-valued’ per se, but it did draw a required performance curve for years into the future. It set a net present value estimate on their future stream of earnings based on some model that both the company and investors believed to be reasonable. And of course this model is based on assumptions – how fast will usage grow, what business model will exist, etc.
Boom! There you have the first reality – the more novel and experimental the model, the higher the risk. And risk brings down valuation, or at least increases the beta, which can reduce the investor pool, thus indirectly bringing down valuation. Let’s be clear – a Twitter that functions as an open platform charging folks for API usage and a services model is riskier than one which makes basic assumptions about consumer engagement and ad revenue. Raising the massive amount of capital at a high valuation had implications broader than just a large deposit into Twitter’s bank account. It fundamentally meant they had a model they were doubling down on, and that model was advertising.
With $8b valuation on its shoulders and the clock ticking for a return to investors, Twitter starts behaving exactly as it should – executing to make the plan a reality. Companies are voracious self-preservation organisms. They eat and grow, eat and grow. The natural state of a large company is to look for new markets, new products. This isn’t a judgment statement, it’s a truism. And it’s also not an inherent critique of the people at the company. In fact, Twitter has historically contained a surprisingly low number of assholes. I have many friends who either work, or have worked, there. Some had great experiences, some have been more mixed, but I’m a big booster of their aggregate talent and motivations. I don’t believe they are ‘evil.’ But the incentive is to grow – to maximize longterm value, which means tradeoffs between users, developers, advertisers, etc.
As Twitter came to grips with that reality, they discovered some nasty realities about the product and business model. First and foremost, with every tweet value potentially leaks out of the system. Or another way to put it, you can create an incredible amount of value for yourself on Twitter with very little friction and without ever paying Twitter a dollar. Organically build a following, tweet out content and their attention/clicks are yours for free. Forever and ongoing. Uh oh.
One way to deal with this leeching is to find other ways to monetize the user coming to your site, if not the organic interaction. And here’s where Twitter encountered two other ‘hmmmms:’
A. The open API meant that a good amount of content was being consumed on non-Twitter owned properties.
B. The nature of tweets meant that very rarely did you need to click over to a Twitter product from a third party experience
So now you had the reality of an unfavorable onsite/offsite traffic distribution which promised to get worse, not better, under the current ruleset. Twitter’s core product experience, which was just recovering momentum after a long period of relative stagnation, didn’t just need to improve, it needed to beat the field. That is to say, it needed to be better than the sum total of what ever other developer could create using the Twitter API. Zoinks!
Okay, maybe I’m being a bit dramatic – Twitter had the brand and capital to ensure the average user experienced Twitter largely through Twitter. And the APIs were not truly open – firehose required a negotiated agreement. But the sentiment is still correct – these were risky realities to build an $8b+ company on. And so they started doing what you might expect: reduce the competition by eliminating certain API use cases and creating more compelling, unique onsite experiences (#Nascar) to help users find value in the core experience, not the ecosystem.
What we are left with is a deeper, but narrower, Twitter. One which will looks more likely to succeed and grow, but in a very specific way. It increases the probability of near term success but reduces the likelihood of emergent behavior. It shifts the responsibility for generating the next 50 amazing features and business models subtly from the ecosystem to the Twitter core team. It’s a bet on the known.
At this point in the analysis I expect some to ask “easy to critique, what would you have actually done?” Tough question especially since I have no inside information and, let me emphasize, TWITTER IS AN AWESOME PRODUCT. I’m so frustrated by Twitter than I use it dozens of times a day 🙂
Now that being said, here are some thoughts. You’ll have to take my word that these aren’t all just in hindsight.
1. Clarity with Developers – Looking Backwards
Developers and technologists are pissed with Twitter because they’ve moved from a warm embrace to hugging them to death. Everyone is culpable – it’s risky to develop on a service that’s still finding its business model, but Twitter was also so aggressively supportive of the ecosystem that one can’t blame 3rd parties for believing they were truly adding value which was appreciated by Twitter management (I have no remorse for devs building ad systems or paid tweet services on the back of Twitter).
And over the past 18 months Twitter just can’t seem to rip the bandaid off. The blog posts are unclear and come from various people in the org. They cause confusion and paralysis in the ecosystem. I love Storify to help me organize tweets to tell a story. Are they in violation of the ToS? I don’t know, and I bet they don’t either.
Lesson learned for entrepreneurs – don’t promise developers stability prematurely. What should Twitter do going forward? Maybe offer a noncommercial API with a limited number of calls per day but fewer restrictions. Allow for hobbyists to tinker.
2. Focus on business models linked to the tweet or the click. Hey Dell, you want links to your store to resolve? That’s going to be .25 per click or 1% affiliate fee. Far fetched? Maybe but I do believe the greatest issue is that Twitter barely participates directly in the economic value they create.
3. Build Better Poweruser Tools (or figure out a 3rd party client licensing model). The 3rd party client ecosystem is going to be really interesting. Clearly they cause problems for Twitter and although it’s a small percentage of users, it likely skews towards powerusers. Since Twitter is a conversation, it’s both a read AND write tool in one. That is to say, if developers were just building better publishing tools, Twitter wouldn’t care. But the consumption and publishing of tweets are inherently linked. So if 3rd party tools are going to persist, Twitter will either need to ignore them or embrace them via a licensing model, ad insertion model, etc.
Whew, so there you go. More Twitter punditry. By the way, in addition to enjoying Twitter, I also committed $100 to Dalton’s app.net. Why?
A. Because Dalton is a friend and I support my friends
B. Because I support crowdfunded technology projects
C. Because I believe app.net could explore the design space with a different set of constraints.
I don’t see this as a zero sum game. In fact, I have the perverse hope that Twitter actually sponsors app.net as a playground from which to fast follow features.
(written quickly on an iPad so will clean up typos later)