Twitter devs: get over it and build real apps
If ever there was a case of the triumph of not paying attention then the storm allegedly brewing between Twitter and its developer ecosystem will make a future masterclass in how not to understand software ecosystems. According to Nicholas Carlson at SAI, this started to erupt when Twitter investor Fred Wilson said:
Much of the early work on the Twitter Platform has been filling holes in the Twitter product. It is the kind of work General Computer was doing in Cambridge in the early 80s. Some of the most popular third party services on Twitter are like that. Mobile clients come to mind. Photo sharing services come to mind. URL shorteners come to mind. Search comes to mind. Twitter really should have had all of that when it launched or it should have built those services right into the Twitter experience.
It doesn't take a rocket scientist to realize that if Wilson is saying that publicly then he's said a lot more privately in his meetings with Twitter execs. Lo and behold...Twitter announces it has acquired Tweetie (mobile client for iPhone) and released a native app for Blackberry (which by the way is very good IMHO.)
It seems this caught a few people by surprise. I'm sitting here and thinking why? OK - so Twitter slightly muffed the Blackberry client announcement by adding in 'official Twitter app' into the blogged announcement. Twitter's Ryan Sarver attempted to paper over the cracks by issuing an email, the essence of which said per SAI:
Twitter shouldn't have referred to its BlackBerry app and Tweetie as its "official" apps, and it will stop using this kind of language. The reason why they created these apps is because new users are confused by how many apps there are. They just want to use Twitter on their iPhone, they don't know or care what "Tweetie" or "TweetDeck" or "Twitterrific" is. Since ending this confusion should increase the number of users who use Twitter, this is also good for Twitter developers, since the ecosystem will get bigger! QED.
Don't be fooled. If they're thinking 'official' then Twitter is already staking out the line between what it wants and what developers will be allowed to pick up. At the moment, much of that might well look like leftovers. In the meantime, Loic LeMeur, CEO Seesmic was caught on the hop. According to Mike Arrington on TechCrunch:
Anyone who didn’t see this coming was in denial. Seesmic
founder Loic Le Meur
...and:
Loic Le Meur yesterday
I have to admit I was not expecting Twitter to step so fast in the mobile client race themselves competing so fast with its ecosystem…As long as we keep moving and innovating and both partners treat each other in a fair way, I think we will all be safe, the hole is big enough and there are many other holes.
At this point I'm going to make three assertions. Any devs out there that are already seething with rage - please read carefully before flaming:
- Developing a Twitter clone is probably one of the easiest things to get off the ground in development terms. I know this because around 21 months ago, a tiny team of rock star devs built the bones of a Twitter for enterprise called ESME. I was part of that team albeit on the non-code side. Even though it was little more than a science project, it was scalable right out the gate and demonstrated immediate business value of the kind Chatterbox is doing today. ESME was partially built on some of the same open source technology that helped Twitter get out of its early embarrassing Fail Whale incidents. We needed a handful of clients to demonstrate how it might look using the web, Air and SAP front ends. It didn't take a huge effort to get wireframes and working clients operational. In fact one client was built on a cross country flight. 3-4 hours. I won't claim these were the polished deal but they were a demonstration that writing code for this class of app was relatively straightforward for the purpose we were trying to demonstrate.
- If a tiny team such as ours could do it then anyone that truly understands the challenges of massively scaleable web apps can have a crack at this. And if you don't own a solid client capable of taking advantage of core application functions then what the heck are you doing in business? It simply does not make any sense. Can you imagine any business application developer delivering server side capability but only a web client of the 'quality' that Twitter offers and hoping for the best? Twitter is now filling in its own holes. It hasn't finished yet.
- Fred Wilson and Twitter have as near as said to the Twitter client (and other) developers: 'We're coming after you.' They realize that owning the client gives them all sorts of opportunities to monetize - initially through ads it seems. I don't understand why people like Loic and others have not seen this. But then I am looking at this through the hawk eyes of enterprise systems where life is rough and competition is fierce in a way that makes much of what we see in the Boy Scout world of consumer apps look tame...and lame. Loic often reminds me that enterprise apps are boring. That may be the case in his eyes but given what is unfolding, he might want to revise that opinion. There's nothing boring - or amusing - about having your lunch eaten.
So what's going to happen? According to Forbes ticked off devs are planning all manner of things not least of which 'might' include:
Two sources deeply involved in the Twitter app world tell us that before Twitter holds its first ever developers conference this week, many developers attend secret meetings where they will discuss an "an open alternative" to Twitter.
"Discussion is not whether to launch an open federated standard," says one source, "but when."
This source tells us these meetings have a code name: Project Shark. "Angry sharks eat big fat #fail whales," he explains.
Another source confirmed plans for these meetings, and told us their purpose is to…
Get a little more coordinated as the opposition. Talk about things they want from Twitter -- mostly more transparency and better access to the API. To talk about an open alternative.
If there is truth in any or all of the above then I'd suggest those same developers hop over to the Apache Incubator and take a peek at ESME's open source iteration. My bet though is this represents little more than saber rattling by some devs whose noses have been put out of joint. Where else are you realistically going to go - at least in the short term? If you think open source can help you then fine. But try explaining that at your next VC pow-wow and hope you don't get flayed alive.
On the other side, Twitter execs have not exactly demonstrated a lot of street fighting savvy so at Chirp I expect they'll go on a charm offensive. Developers with antenna on high alert will be watching for any signs of BS. But if Fred Wilson and the other VCs have anything to say about the next step (and they do) then Twitter might as well make acquisition choices for the desktop client, a URL shortener and photosharing uploader now. Inflict the coming pain now, lay out the future turf and let's all move on.
In the meantime, those same developers should take heed of what Wilson is saying:
...it is clear that you can build large businesses on top of a social platform like Facebook and Twitter. And because Twitter is so open and so lightweight, I am surprised that there aren't more "new kinds of killer apps" to quote my friend who I started this post with.
Fred may be surprised and Loic may not like it but Twitter devs have got to understand that we're talking about b-u-s-i-n-e-s-s and that some of the best cues come from the enterprise experience. Here's an analogy. Most current business applications are built around a core of good ol' fashioned book-keeping. Even now and often after 10, 15 or 20 years, those core apps are still being developed in some fashion. Those vendors that have developed successful ecosystems are open in much (but not quite) the same way as Twitter. Here's the difference. Successful developers have added something the core app owner cannot afford to devote resources towards but which meet a profitable niche. StockTwits anyone? That's an example Wilson quoted.
Wilson laid out four broad areas where he encouraged developers to turn their attention. How much clearer could he be? They're more demanding than 'yet another URL shortener' but likely to deliver back far more in enterprise and stakeholder value. If that message has not sunk in then many of those that have been merrily playing around with Twitter's open API are in for a shock to the system.
UPDATE: Robert Scoble provides a useful update on Loic's response to the potential Twitter 'problem.' It's worth the reading.