Editor’s note: In this guest post, serial entrepreneur Nova Spivack gives Twitter some suggestions for how to make money. Spivack’s latest startup, Bottlenose, is looking at new ways to mine the Tweet stream.
I’ve been puzzling over Twitter’s recent tactical moves around their API, Ubermedia and Tweetdeck, for a few months now, and it just doesn’t add up. In fact I think Twitter’s current strategy may take them in a direction where they end up missing out on their biggest potential win.
If Twitter continues to go down the media company path, without incorporating their API into the plan, that could not only force a large part of their ecosystem to go elsewhere, but it could deprive them of a much larger potential infrastructure revenue opportunity, and could even end up costing them the company.
After all, Silicon Valley is littered with the burned out wreckage of once-great media companies that failed create and keep third-party app ecosystems: AOL, Friendster, MySpace, Yahoo – to name a few. It’s very hard to maintain leadership as an online media company without an ecosystem of outside apps increasing reach, innovation, and stickiness.
In light of this, I’ve been exploring an alternate path for Twitter that leverages their API in a much bigger way, and this path appears to be a better strategy. According to my own experimental revenue projections for Twitter, this alternative path is not only a good tactical move, but it’s a good business move because it increases Twitter’s reach, number of active users, and revenues massively.
This path fulfills the promise of Twitter as an infrastructure, without sacrificing the media company play. A media company + an infrastructure is a much stronger strategic position to be in than either on their own.
Another side-effect of this proposal is that it eliminates the need for Twitter to buy Tweetdeck, or Ubermedia. It makes the wholediscussion about the risk of Tweetdeck and Ubermedia to Twitter completely irrelevant, a non-issue, and will save Twitter $50 million in unnecessary acquisition costs.
It also eliminates the tension between Twitter and their ecosystem of third-party client apps. And it returns more revenues to everyone, especially Twitter. In the end, this could make Twitter a much bigger and more important company, and would certainly lock in their dominance of global realtime messaging and advertising.
To understand my proposal, first, what is Twitter really? Well, if history is any indication, it’s a messaging infrastructure for the Web. Let’s shelve the question of whether it’s the optimal messaging infrasture (it’s not, by a long-shot), but it works well enough for the moment.
Twitter’s APIs are a big piece of how Twitter grew so quickly: Twitter surged because of third-party developers pumping data in and out of Twitter via these APIs in all manner of apps and services, which massively extend the reach, innovation, and impact of Twitter.
Instead of abandoning their DNA and clamping down on API use and competing with their own ecosystem, my analysis shows that Twitter would do far better through a combined strategy.
In the combined strategy Twitter would continue to have a destination portal and their own official apps, but would also actively encourage – and monetize – an ecosystem of third-party apps on their APIs, including client apps that effectively competed with their destination. This competition would however not harm Twitter, it would make the ecosystem even bigger, and would deliver very significant incremental revenues to Twitter as well.
The key to the combined strategy is a new way for Twitter to monetize their API’s. Let’s call this the “freemium API” option. Here’s how it works conceptually:
Twitter would change their API terms to give third-party apps two choices: Either use the API for free but accept in-stream ads from Twitter, or pay a very nominal fee per tweet (around $0.1 per thousand tweets in or out of the API, a 10 cent CPM). Apps that opt to pay for the premium API could easily monetize with their own ads or subscriptions to more than compensate for the 10 cent CPM to Twitter, and would make money on the delta. Even if you think that’s too high, Twitter could cut that in half and still make money.
Here’s the model in a little more detail:
- Third-party apps that don’t mind carrying Twitter’s ads could use the free API. They would be able to run their own ads outside the stream, but not inside the stream in their apps – only Twitter’s ads could appear inside the stream for the free API. These ads would come from Twitter and could even be personalized or targeted per user or topic.
- Third-party apps that either don’t want ads at all, or don’t want Twitter’s ads, could use the premium API, pay the fee, monetize, and make money on the spread. They could monetize with their own ads or through subscription models or commerce or whatever they want. In this option, third-party apps would not be allowed to inject their own ads into the outgoing stream; instead they could display their own ads interleaved within the stream, in their user-interfaces, but these ads would not be pushed out to Twitter, they would only appear for their own users. This way Twitter would not be flooded with ads.
By launching this freemium API model, in addition to their existing portal business and their official client apps, Twitter would be able to monetize their entire ecosystem, including every third-party app. The beauty is: Twitter gets paid no matter where a user enters their network or views content; Twitter makes money from 100% of all tweets and views. It’s a vastly more scalable business model than just being a destination or media company and trying to own 100% of the user-experience.
As an experiment, I’ve run the numbers and they look good; See for yourself. Of course these numbers are based on anecdotal data, such as a rumor I heard from a credible source that Twitter’s actual revenues this year are closer to $75mm. It could add another $20 million off the bat with a freemium API strategy.
My projections simplify matters in several dimensions for the sake of convenience in sketching out the scenarios, and perhaps have growth rate and audience share assumptions that are debatable – but regardless, even if we were to tweak the model a bit, the conclusion is the same: Twitter would have a bigger audience and greater revenue growth if they included the Freemium API model and made it a priority.
Interestingly, Twitter currently licenses all of its bulk data through a third-party company, Gnip. Gnip prices their data at $0.0001/tweet or $0.1/1K tweets – exactly what I proposed in my model. Instead of that money going to Twitter, some or all of it is presently going to Gnip. This makes very little sense to me.
Why would Twitter give away their API – their platform – to an outside company, especially when at its root Twitter is an API? I think it would ultimately make more sense to take that in-house, and if I were Gnip I would be worried about that. Perhaps Gnip is an acquisition target by Twitter in the future? In fact, Gnip is a much bigger potential threat for a company like Twitter than Tweetdeck or Ubermedia are, in my opinion.
So far we’ve analyzed what happens if Twitter DOES take the strategy of offering a freemium API. But what happens if they DON’T? Either of two sub-optimal outcomes:
- If Twitter allows 3rd party clients but does not monetize the API in any way – then eventually 3rd party clients will take significant market share away from them. This is the problem they are facing with Tweetdeck and Ubermedia currently.
- If Twitter tries to stop (A) by blocking or clamping down on 3rd party clients – it won’t work. First of all this will cause existing 3rd party client apps to leave the Twitter network, taking large portions of high-value power-users with them. There are numerous stealth projects now underway to create alternative networks to Twitter, and sooner or later one of these will succeed. More importantly, if Twitter blocks use of their API they will cut themselves off from being a platform and infrastructure, making them vulnerable to attack by competing services (like Facebook or Google) that might be more developer friendly and that take more of a platform approach.
The conclusion of this is that it is clear, to me at least, that if Twitter turns its back on their platform and API DNA, they are missing out on what may be their most important tactical opportunity.
Being a platform and having thousands of 3rd party apps will increase their reach, massively increase adoption and engagement, and create a much more powerful and sticky network-effect. In short, killing their own ecosystem to save their portal business would be cutting off their nose to save their face.
So what should Twitter do? Simple. They should not buy Tweetdeck or Ubermedia. There is no need to worry about Ubermedia or anyone else. They should not clamp down on their API or try to block third-party client apps.
Twitter could solve all these problems, and double the value of their business, in an instant by simply launching a freemium API, along the lines of what I’ve proposed here.
If Twitter simply embraced their API roots instead of turning against them, all their “frenemies” would become friends again, and Twitter could focus on building the best realtime ad network and messaging infrastructure in the world, instead of competing with their own channel partners.
If Twitter doesn’t do this, then mark my words, they will eventually lose their dominant role, as well as all the goodwill they currently have. And they will force the market to come up with competing solutions.
At the end of the day, without an ecosystem, Twitter’s network effect will fall apart pretty quickly. If Twitter loses their ecosystem by competing with it, they will end up in the graveyard of once-great Internet companies. I personally would not like to see that happen.
I would like to see Twitter function as an infrastructure, not merely a media company. It’s better for Twitter, it’s better for their ecosystem, and it’s better for the world. But if they fail to do that, I’ll happily embrace better solutions when they emerge.