Twitter turns five today. The company has grown at a rate that most can only envy. Critical to that growth has been the development of an ecosystem around it. But Twitter’s relationship with its ecosystem has always been a bit shaky.
Concerns about the future of the Twitter ecosystem began last spring when Twitter acquired Tweetie and began to launch its own mobile applications. Later, Twitter enhanced its website in an effort to drive more usage through the site, as opposed to third party apps. A month ago, third party Twitter apps UberTwitter and Twitdroyd had their access to the Twitter API suspended as a result of a trademark debate and other “policy violations”.
Last week, the debate hit a tipping point when Platform lead Ryan Sarver shared that developers should not “build client apps that mimic or reproduce the mainstream Twitter consumer client experience”. In other words, don’t bother building new client apps like Tweetdeck or UberTwitter.
Ecosystems are critical for the growth of platforms. Google is the obvious example, but Salesforce.com has its Dreamforce program, Facebook has cultivated an ecosystem, while traditional software companies like Microsoft or SAP could not have grown as quickly without the systems integrators and third party applications that sprung up around them. But ecosystems cannot thrive on their own. As in the biological world, they require nurturing and cultivation. For technology companies, that means ongoing communications and behavior consistent with that message.
As a developer, becoming part of the ecosystem has its benefits but also many negatives. The obvious risk is that your success is tied to that of the underlying platform. Beyond that, you run the risk that the platform will see your functionality as a critical part of the core, and embed those capabilities into their own applications or APIs. Of course, the platform could even shut off your API access, or more subtly, remove data you depend on from those APIs.
Investors are even more cautious, wanting to be certain they are funding a real company, and not simply a cool feature that will eventually be embedded into the platform.
For the platform, cultivating an ecosystem is not easy. Inside the company there are always many arguments against it. It often seems like a zero-sum game, where every dollar (or ad impression) that goes to a partner is one that the platform does not receive. Developers often believe they know what’s best for the user and that third party apps may not present the underlying platform as well as they would. But a thriving ecosystem helps drive adoption and creates much greater barriers to entry for others. And while your platform developers may have some ideas on how people will use the platform, it’s much more likely that other use cases will spring forth from having many 3rd party apps.
So, what does Twitter need to do?
First, it needs to build a stronger relationship with the developers in its ecosystem. And a starting point would be to have a single point-person focused on that. When I first tweeted the idea that Twitter needed an ombudsman, I got an @reply from Dick Costolo, Twitter COO. Dick’s a great guy and I was thrilled to know that he was listening, but he is clearly focused on the advertising side of the business and has enough fallout from the so-called #dickbar that he can’t be the key contact. Ryan Sarver owns the APIs and is a great technologist, but he’s not focused on communications. I think they would be well-served having a point-person focused on the ecosystem, much in the way that Matt Cutts is the key Google spokesperson on SEO-related issues.
The next key thing would be to meet with some of the developers and based on those conversations, share the basic philosophy of how Twitter plans to work with the members of its ecosystem.
A platform has a lot of flexibility in determining how they want their ecosystem to grow. They can have a very open approach or a more closed one as Apple does. But the key is strong and consistent communications, no surprises and keeping the API open so that new features available on the core platform are all exposed through the API.
I’m optimistic about Twitter’s future. I don’t see the latest ripples as a death knell, as Alexandra Samuel suggested in her well-thought out HBR post last week. But I think it critical that Twitter begin to engage its developer community on a much more active level to ensure that success.