Apple has slowly been making changes to improve how people search and navigate the App Store. Part of the way it does that is via small acquisitions — some of which stay under the radar for years.
TechCrunch has learned that Apple quietly bought a startup called Ottocat some time ago, which had developed a system to organize and surface apps on the app store based on “nested” categories of increasing specificity. A version of that system now powers the “explore” tab in Apple’s App Store.
There is precious little evidence of the connection between Ottocat and Apple — no LinkedIn employment changes, no announcements to Ottocat users — save for one thing. One of Ottocat’s co-founders, Edwin Cooper, authored a patent that was granted to Apple as the original assignee. It looks like that patent was filed by Cooper as an employee of Apple. That patent, for a “System and Method for Divisive Textual Clustering by Label Selection Using Variant-Weighted TFIDF,” pertains to the kind of technology used both by Ottocat and Apple’s explore feature.
The acquisition looks like it may have happened some time in 2013. It was in October of that year that Ottocat’s website went down with the short message: “Ottocat is no longer available.” Now, pointing your browser to Ottocat’s URL gives you a blank page. Ottocat had a very short life: it only opened its beta to the public in May 2013 after developing a working prototype in January 2013.
The timing of Ottocat going dark in late 2013 links up with Apple announcing “explore” in the App Store in mid-2014, part of a series of updates made public during the WWDC that year to improve how apps can be discovered in the increasingly large and unwieldy App Store catalog.
So what was Ottocat?
We never covered Ottocat in its brief life (others did), but in a nutshell, its technology essentially addressed pain points on both sides of the App Store: for users unable to find specific enough results for subject-based app searches when they don’t have a specific app in mind; and for developers unhappy with how well their apps could be discovered among a sea of 1 million+ other apps.
The premise was to do away with keywords by categorizing apps into increasingly more specific subcategories that worked on a “drill-down” principle — eliminating the guesswork and potential inaccuracy of keywords altogether.
Or, as Cooper’s patent describes it:
In a world where billions of digitized documents exist, technologies that make large sets of documents more tractable are in significant demand. One such area of technology is the search engine. By submitting a query, a user may restrict the set of available documents to those most relevant to the query. Another such technological area is that of document clustering; by automatically creating groups (clusters’) within a large set of documents, that large set of documents can be divided into consumable parts, and thereby made tractable to a given user or set of users. Given a hierarchy of document clusters, each with a descriptive label, a user may browse quickly to those subject areas of greatest interest, and may further refine the chosen subject area through choosing an appropriate subcategory. Assuming that the clustering technique is applied recursively until no more that a pre-specified number of documents exist in a cluster, the user may browse document to an arbitrary degree of specificity in terms of clustered subject area.
For example, rather than searching on “guitar” or scrolling through the full selection of music apps that the term might call up, or the chart for the most popular music apps — which can contain streaming apps, apps that are designed to work with specific hardware, apps that let people use their phones to play music, apps that teach them how to play a specific instrument, and so on — you can start to look at specific subcategories to find a selection of apps you may want to download.
This is what such a search looked like for interactive children’s books:
For those of you who use the “explore” feature on the App Store, the basic idea of how all this works will sound familiar.
Ottocat, which claimed to index the entirety of the App Store in this way, also included extra details such as the average star ratings, how popular the app has been, and when it was last updated.
Ottocat was co-founded by Edwin Cooper and Michelle Cooper, both repeat entrepreneurs, in 2012. Michelle holds a PhD in molecular biology and has extensive experience in marketing and fundraising.
Edwin is a computer scientist who once noted on his LinkedIn profile that he is interested in, among other things, data science, natural language processing, machine learning, mobile devices, “and the fallout from Big Data.”
We are not sure if said Coopers are at Apple now. If they are, it’s enticing to wonder what they might be working on, given the bigger decisions Apple may be making around search in other products like Safari and its mobile and desktop operating systems.
While Yahoo and Microsoft are reportedly duking it out to take over Google’s place as the default search in Apple’s Safari browser in the U.S., we’ve heard that it might be just as likely that Apple might try to bring more of search tech in-house. This would be in keeping with the massive effort that Apple is putting in to making other services like maps into a native experience. It would also play directly into the kind of expertise Edwin Cooper has amassed.
Interestingly, this would not be Edwin Cooper’s first bite of Apple. In 2011, he sold a previous startup he founded to Oracle. That company, InQuira, developed a CRM platform for third parties to offer self-help to its users. By coincidence, the search feature on Apple’s support site is one of the services powered by InQuira’s technology.
We have reached out to Apple and Ottocat’s co-founders for comment and more detail. We will update as we learn more.