Lots of interesting news this week (and still a day to go), but perhaps the most interesting bit is Twitter's Fabric launch.
Everyone interested in this topic should read @honan's piece on it here http://www.wired.com/2014/10/twitter-fabric-sdk/, but I have a slightly different perspective on the strategic importance of it. Perhaps colored by the fact that we're developers.
Ultimately I don't think it's right to think of this as a new effort to identify developers as a potential market and fabric as a new set of services offered to them that's completely discrete from consumers/tweet (the way, for example, AWS is discrete from consumer retail ecommerce).
Rather, this has to be thought of as directly adjacent and highly relevant to what twitter is, which may be expressed to the market more or less as a media company, but is, in actually, fundamentally an interest and attention graph.
By providing a tool that can listen in as users are registering for new apps, selecting friends to connect with in apps, and logging in and out of apps, Twitter gets data that is highly valuable and highly additive to what they already know about a) who & what people follow & pay attention to; b) who people communicate with; c) what topics and apps they may be likely to follow and/or install.
Why is this interesting? Well, let's start with their arch-competitor, facebook. Facebook doesn't break out mobile install ads specifically, but 62% of their ad revenue comes from mobile, and we know a good chunk of that is mobile install ads -- and anecdotally we know that in addition to being the go-to source of paid installs for mobile developers and large brands & retailers, there is an entire ecosystem of "preferred marketing developers" being built on mobile install ads. In short, lots of money flowing through that system. And, facebook has a huge advantage on targeting those ads (ie driving their performance, and maximizing their revenue-vs-inventory) because of the number of developers that implement sign-in with facebook, and the number of users who use it.
If I know who is using an app, who they're connected to, what other apps they're using (because of my omniscient view of all the apps a user connects to facebook with), who their friends are and what apps they're using, etc., I have a difficult-to-unseat advantage in competing for ad dollars (akin to Google's lock on search, ads and publisher footprint in the heyday of adsense for the web).
Back to Twitter: with that in mind, I believe Fabric -- and Digits in particular -- is their play to leapfrog facebook and get at an even more fundamental graph, the phone directory. (As @honan pointed out, Fabric lives in an SDK, like facebook's mobile tools, which means it can see lots of things other than signins, and it can listen explicitly to many other events over time.)
I've long believed that Twitter's biggest unexploited asset was its login/friendfind-with-Twitter capability, which should have helped them know what apps a user would be most likely to install, and build a massive pay-per-install network on top of this. (In this capacity, attempts to discourage developers from using the twitter API to extract the graph for this purpose were just stupid.) They've taken some decent steps toward this, and we and lots of developers we know are using, or at least testing, mobile install ad formats on twitter. But it is not the distribution and revenue juggernaut it could be -- and this is a huge step in that direction.
Given all the trends moving away from facebook these days (privacy pushback, declining user interest), combined with all the interest in contacts as a graph, Twitter's timing could not be better.
As for users? Eh, what's one more massive web player trying to suck up all your data?