/ app.net

Tweetbot, and Why My Days on Twitter are Numbered

Today, I purchased Tweetbot for Mac. I paid $20 for this piece of software that will do nothing but serve as a avenue for me to tweet my thoughts for the days that I remain on Twitter – and those days are numbered.

You might say, “Matt, that seems silly. Why would you spend that much money on a Twitter program? That’s a waste!” Indeed, under other circumstances I might agree with you. However, the fact of the matter is that the author of Tweetbot, Tapbots, is being fundamentally screwed by Twitter. In fact, every company that sought to build a product line off of the ecosystem that Twitter created is being screwed right now, and Twitter doesn’t much seem to care.

In August, Twitter imposed new restrictions on applications that meet the need of a standard Twitter client (you can read the blog post here). Specifically, their new restrictions function as follows:

Additionally, if you are building a Twitter client application that is accessing the home timeline, account settings or direct messages API endpoints (typically used by traditional client applications) or are using our User Streams product, you will need our permission if your application will require more than 100,000 individual user tokens.

Now, I’ll do a little bit of translation for those of you who are unfamiliar with how Twitter’s authentication system works. Essentially, one user token is equal to one application (e.g. Tweetbot) being able to access one Twitter account and being able to do something as that user. So, if you (as a company) are building a traditional twitter client, Twitter has essentially imposed a 100,000 user account limit on the number of user accounts your application can access across all of your customers. For added oomph, there are some users (like myself) who have more than one Twitter account. That means more than one token.

The truth is that 100,000 is a pretty large number regardless, but the result of this large number meant an increased price point for Tweetbot: $20. There’s just no way around the fact that Tapbots does excellent work. I’ve honestly been spoiled by their apps. They’re beautiful, featureful, and an all around pleasure to use. They’re everything an application should be. Honestly, if it weren’t Tapbots we were talking about, I probably wouldn’t have shelled out $20 for Tweetbot for Mac, but I can’t deny that they do good work and are worth the coin. I want to do everything I can to support them, including paying a bit more than I might otherwise for a Twitter client.

To add insult to injury, Twitter has done very little with their own desktop client in quite a long time. In fact, I don’t think it’s received any substantial updates since Twitter acquired atebits and rebranded Tweetie as Twitter for Mac. Since then, the guy behind the client they acquired has hit the reboot button on atebits and is doing his own thing, meaning, I suspect, that we shouldn’t expect any updates to Twitter for Mac anytime soon. Twitter’s neglect of their desktop client has resulted in it being the worst in the arena. It doesn’t support read later services like Readability, doesn’t provide any selection in terms of link shortening or image uploading, doesn’t fully support Instagram links in the client (you have to open your browser to see them), and has a lot of weird bugs that have never been addressed. Their mobile apps are marginally better, but not by much.

So, let’s review: Presumably aware of the fact that they have desktop and mobile client applications that are the functional equivalent of quadriplegic assassins compared to their competitors, Twitter decides to place a hard user token limit on all these competing applications that deliver a superior user experience and feature set.

My thoughts exactly, David Tennant.

The only reasoning I can come up with for Twitter’s recent behavior is that the people placing ads on their service wanted the company to have more control over the client ecosystem to guarantee that their ads will display. What I *don’t *understand is why Twitter didn’t take any one of the one hundred more reasonable avenues toward meeting that goal. For example, requiring that traditional client apps display promoted tweets. Oh! Perhaps charging traditional client apps a fee per user (not per user token) to use the API to recoup the cost of the lost advertisements?

Whatever the reasoning behind this debacle, this decision by Twitter is the most developer-hostile policy that I’ve seen from a social networking company since… well… social networking became a thing. To add to the irony, Twitter owes its success, at least in part, to the vibrant ecosystem of apps that developed around the service, unlike Facebook that developed independent of such apps. Of course, I’m not saying anything that everyone doesn’t already know. I’m just partially in shock that they would make a play this… brain dead.

For this reason, I suspect my days on Twitter are numbered. Recently, a new service named App.net started up. Functionally speaking, they are identical to Twitter. The only noticeable difference from a user perspective is that App.net is a paid service – either $5/mo or $36/yr. Honestly, I foresee myself making the jump sometime soon. I’ll probably straddle the fence for a bit, or setup an IFTTT hook such that App.net posts syndicate to the corresponding Twitter accounts, but to be honest – if Twitter had implemented a pay scheme to use apps instead of taking harmful action against developers, I probably would have paid the toll. It’d be worth $36/yr to be able to continue to use Tweetbot. Yes, in addition to the cost of the actual client itself. Alas, if only, if only.

I suspect we’re going to see a few things happen over the coming months. The first one is probably going to be a brain drain on Twitter. People like me, who don’t take kindly to companies biting the developers that feed them, will leave the service in due course in favor of services like App.net. Either this will result in a wake-up call at Twitter, or it won’t. If not, the inevitable will happen. Everyone except the high schoolers will eventually migrate to App.net or similar services, and Twitter will slowly go down the same toilet that MySpace and others have gone down. I certainly hope that doesn’t happen, but having so little business sense and being hostile to the group of people who were your early adopters and evangelists means you’re going to eventually be left with a lower quality user base than you had when you started. This is how companies slowly rot into irrelevance.

Don’t believe me? Compare the Twitter activity of Paul Haddad, one of the Tweetbot engineers, both before and after the announcement about API tokens. Now compare his app.net activity. Also note that Tapbots has released an App.net client for iPhone and iPad, and the hypothetical discussions happening about an App.net client being developed by Tapbots.

Ball is in your court Twitter.