A few months ago I left a pretty well paying job at Zend to start developing mobile applications. I still do some work for Zend, but most of what I do during the course of the week is building apps for mobile devices on my own and writing music. The first one I started working on is called SkwawkIt. I started it about two months ago and have been working on it for a while. It was based off of some problems I had at Zend getting people to retweet important tweets.
Then I started hearing about changes that Twitter was going to make to their developer rules. After contemplating it for a while I took the step of removing about 2/3 of the functionality. Why? Because what I was hearing was that Twitter wasn’t going to allow applications that competed with their own mobile applications. There were some parts of the app that were acting like a Twitter client and since I am a very small shop who only had a good idea, I did not want to spend several months developing an application only to have Twitter say “nuh uh”.
Then the new rules of the road came out this past week and it turns out that they were worse than I had feared. Not only were they going to be limiting the types of apps, but also API call rates and the like. Because of that I looked at what I was building and have become nervous with even the 1/3 of the app that I was leaving in there. So once again I had to re-evaluate what SkwawkIt was going to look like and if I could keep on going. Considering I had already spent two months working on it I decided to keep it going, but to remove some more features so-as not to run foul of what could be perceived as infractions against the rules of the road. Perhaps the changes aren’t as bad as I fear, but Twitter has introduced a significant amount of uncertainty into their ecosystem. And uncertainty is bad when you are trying to build a business.
So what Twitter is now doing, to their detriment, IMHO, is limiting what third parties can do with their API. What they are basically saying is “quit providing cool ways of interacting with Twitter and build analytics tooling instead.” Poo. Control, that’s what it’s really about.
But how did Twitter become so successful? Were they in control of their success? μὴ γένοιτο, as my Greek professor used to say. Literally, “May it never be!” in the strongest possible terms. It was the third party developers who made use of the Twitter API that made Twitter successful. Those 80% of tweets that originate on Twitter infrastructure would not occur without clients that make it easy to interact with Twitter. For example, I would not be using Twitter if it were not for Tweetdeck.
Why were these third party developers successful? Because Twitter, even if they have the highest genius to employee ratio in the world cannot scratch every itch and solve every problem. And these third party developers started making money by doing that. So what does Twitter do now? They limit them, rather than setting them free.
And this has made a lot of developers mad or, like me, frustrated. Rather than limit me, here is a radical suggestion.
Charge me for usage!
I like having free access to Twitter, but if I am building a paid service that makes Twitter easier to use or solves a problem in some unique way, then it simply a cost of doing business.
For example, if I have a service that someone pays $10 a year for to manage their tweets (this is just a random number) why would I not be willing to pay Twitter 50 cents to be able to post as needed? While 50 cents may not sound like a lot I hear that Twitter has around 1 million apps. Most of those apps are probably not in use, so let’s assume for the sake of argument 25% of them are in use. Let’s also assume that each app has only a thousand users, on average. Given these assumptions, 50 cents per user per app per year comes out to $125 million. That’s not exactly chump change. With Twitter’s current RotR changes they get zero. Which is better?
When it comes to a pie, Twitter seems to be looking for a larger slice of a smaller pie. I would argue that it is in their interest to make the pie as big as possible. The best way to do that is to release the creative juices of hundreds of thousands of developers who use Twitter. Don’t limit us, unleash us! Rather than make us worry about whether you’re going to be a dick about something just charge us for access to the service and let us worry about our own bottom lines. I don’t mind paying some kind of a usage fee if I am making money off of Twitter.
In other words, Twitter, you worry about the infrastructure, API features and basic usage. Let the third party developers figure out how to monetize the system and in turn pay you for the access. You may get less as a percentage but you will have a larger pie and a healthier market.
Dittos for Facebook. You have a problem with monetizing mobile? Provide your own client for basic functionality, but create the infrastructure for creating engaging mobile applications and charge for access to it. I, personally, have a very hard time understanding this fascination with command and control structures in social media. Social media properties work specifically because they are not command and control.
And if the CnC fails, so does the whole thing. Why not introduce competition within your ecosystem and let the best solution to a given problem win? Then your developers are happy, your consumers are happy and you have income that does not rely on advertising. Facebook ads are largely useless and Twitter ads are horribly expensive. With Facebook’s recent stock freefall we are seeing the limitations of advertising as a way to support a social media platform. Twitter is no different.
A social media platform’s user’s are its greatest asset. And people are willing to pay for access to that asset.
It’s a lot less creepy than the analytics that are done to serve up ads, and it keeps in spirit with what social media is all about. Namely that it is the social network, not the backing company, that provides a) the interaction, and b) the content.
Besides simply charging for access is there anything else that Twitter can do? Yes, IMHO. That will be another blog post.