The climate for third-party Twitter apps has been wintery cold of late. A year ago, Twitter raised questions about the future of outside clients, flat-out telling developers to stop making them. Then, just a couple of weeks ago, Twitter put out another ominous-sounding post that caused another wave of worry.
At the heart of most of the susurrus is the fact that nerds love their third-party Twitter apps. Twitterrific, Tweetbot, Echofon and other apps in their arena have become beloved tools that many hardcore Twitter users would be loath to give up.
Twitter is clearly moving towards offering a user experience across all of its platforms that allows it to offer advertisers new ways to entice users to click. This includes their timeline ads as well as products like expanded tweets. Because third-party clients don’t offer these features, at least as of yet, there is concern among many heavy Twitter users that they will outlaw third-party clients once and for all, leaving us with only Twitters first-party clients to use.
Note that when I say ‘clients’ I mean apps that give you the Twitter ‘experience’ in similar ways to the official Twitter website, not apps that integrate the Twitter API into their apps. Twitter is just fine with those apps and encourages people to use its API to build interesting new products that leverage it.
That’s why the question of how many people really use third-party clients to access Twitter in a traditional way is an interesting one. If Twitter were to shut off third-party clients’ access to its API, could it get away with it?
Developer Benjamin Mayo has taken an interest in the topic and decided to collect a sampling of Tweets and break them down by client, in order to determine just how much ‘shutting off’ standard third-party clients would ‘hurt’ for Twitter.
To do this, he recorded a sampling of 1 million Tweets across a 9 hour period, collating the clients that were used to post them, in an effort to track down just how many tweets are posted from apps outside Twitter’s own ecosystem.
Note: We verified Mayo’s results against those provided to us by popular Twitter developers in order to ensure that they represent an accurate estimate. As far as these developers told us, Mayo’s numbers are very similar to what they’ve seen. The sampling isn’t perfect, as it doesn’t take into account every time zone and wasn’t taken from Twitter’s highly controlled firehose, but it’s a good basic estimate.
Here’s a listing of the top 20 posting entities, Mayo has the full list in his post:
|Twitter for iPhone||152116||Yes|
|Twitter for Android||116765||Yes|
|Twitter for Blackberry||115244||Yes|
|UberSocial for Blackberry||34502||Yes|
|TweetCaster for Android||11746||Yes|
|Twitter for iPad||9603||Yes|
|Tweetbot for iOS||6845||Yes|
So, what were his findings? After normalizing the data and tweaking it to weed out those results that were ‘Tweet buttons’ or other non-client entities, he ended up with a set of 291,899 tweets across 118 top posting tools to analyze.
Out of those tweets, he found that first-party apps accounted for over 77% of all tweets from clients, with third-party apps ending up with only 23%.
These numbers, as Mayo points out, are good news for Twitter if it decides to remove access to third-party clients, as a large number of those people using them would likely switch to a first-party client without too much grief.
It’s bad news for those of us who love our third-party clients, though. And bad news for the developers who make those clients. If Twitter feels that they could make the change without taking a big hit on usage, they might very well do it to provide the ‘consistent’ experience that they talked about in the post mentioned earlier. That experience would allow them to be more effective to advertisers and consolidate that extra 23% into users who were getting ‘the official message’.
There is no evidence that Twitter is going to shut off third-party clients completely, at least for now. I haven’t heard anything concrete besides the fact that changes are coming and that third-party developers are very, very nervous about the future of clients. Maybe Twitter will allow those already existing to stick around, but not allow new ones, or perhaps it will outlaw clients entirely.
Either way, it doesn’t look like it will hurt Twitter that much.
Mayo has done a lot more analysis of the numbers and explains his methodology fully in his post, I encourage you to check it out here.
P.S. Did you see those numbers for BlackBerry? Wow!