Quettra Portrait Taps Your ‘App Graph’ To Power Personalization

Two of the most-used metrics these days that gauge how well an app is doing are downloads and registered users — numbers that can sometimes reach into the millions and point to a degree of popularity. But what downloads don’t indicate is how to improve retention, or getting users not just to download an app, but to open and use it once, twice and ever more regularly.

Quettra, a mobile analytics startup backed by the likes of Google Ventures, CrunchFund (founded by TC founder Michael Arrington) and Horizons Ventures among others, is today coming out of stealth mode with a product that it believes could help solve that by helping developers fine-tune their apps more specifically to users initially interested enough to open those apps in the first place.

Portrait, as it is called, is a new twist on app and user analytics that is at once very light-touch and also potentially very revealing.

Today there are dozens of things that can be tapped to reveal more about you as a mobile user — sensors on your mobile device that measure for example physical activity, location and even barometric conditions; what you browse on the mobile web; what you click on in specific apps; how much data you use; what phone you own and so on.

For Portrait, Quettra is effectively tapping only one thing: your “app graph” — or a scan of what apps you have already installed on your iOS or Android device. Using that information, Quettra has developed a larger big data play that provides some intelligence around what your personal app catalog says about you.

That information, in turn, can be used by developers in a few different ways: to help tailor the onboarding process for a new user (say, with more specific suggestions of content, or ways of using a product that fits what you might want to do, or your possible ability); to provide more relevant notifications after initial use (not a bad idea considering that we are fast hurtling into over-notification oblivion); and to help provide in-app ads that might actually mean something to the person viewing it.

Screenshot 2015-05-15 at 5.35.42 PM


Quettra’s founder, Ankit Jain, the former head of search and discovery for Google Play, says that the intelligence database that gets matched against a user’s particular catalog of apps is already 125 million profiles strong, in part because a private beta of Portrait has already been trialed by a number of publishers.

They include Sungy Mobile (who is another investor in the startup); Haptik, App In The Air, Echo Lockscreen, Dil Mil, RV App Studios and Junglee Games. (There are others that Quettra is not naming.)

Developers implement Portrait analytics by way of an SDK, and then when a person opens up an app with it implemented, that app can make an instant scan of the device, which is then fed back to the developers with accompanying analytics and to start pushing better content their way.

It’s also integrated with marketing automation platforms like Parse, Urban Airship and MixPanel, as well as ad networks like Admob, InMobi and MoPub.

When Quettra first announced its funding while still in stealth, one of the spins about what the company did that it was all about advertisers knowing even more about you than they already do. Perhaps because of this, or perhaps because Quettra is simply just revealing more about what its aim is, this isn’t what Jain emphasizes much today. Users completely opt out of having their app profiles known (on iOS and Android it’s part of the privacy settings that also manage how ad networks can track users). In those cases, users are completely anonymous and just get generic, non specific profiles via the Quettra SDK.

But more importantly, in those cases where people are not opting out of these options, Jain believes that what Quettra offers is a way of democratizing what kinds of tools average developers (and even larger ones) have at their disposal, putting their ability to personalise content on par with the kind of work that companies like Google are already able to do.

“The mobile world has really focused on app downloads and acquiring users, but I think  the rhetoric of retention is going to increase in coming years,” he says. “We want to lead that charge.”

He notes that companies like Google, Facebook and Twitter are all already going down this road. “They are all creating interesting datasets,” he says, something that Jain with his past experience at Google knows something about. “It’s all about access to understanding users.”

Quettra is breaking some new ground here, but all the same, it’s not the only one that is looking at ways of creating products that help personalise apps. AppBoy is working on ways of improving smart notifications, while another startup called Sensiya is also offering an SDK to developers to get a better idea of the preferences of their individual users.

Jain says Quettra has filed several patents for its technology and how it uses it to generate its profiles. And the fact that there is a specific product out called “Portrait” implies that there will be some other services, using perhaps other techniques, down the line for Quettra, too.