Semantic App Search Engine Quixey Now Digs Into Apps Themselves To Serve You Direct Results

Next Story

Mobile Messaging App Line Claims 2M New Users In 24-Hour Period After WhatsApp Outage

Quixey — the semantic app search engine that has raised some $74.2 million from the likes of Alibaba, Eric Schmidt’s Innovation Endeavors and others — is raising the bar again in its bid to make apps and the data that they hold more discoverable. Today the startup is launching a new feature that it calls “Functional Search” — which will let users not only connect with the right apps by keywords and phrases, but with content from within those apps that answers a user’s specific question.

In other words, you can use Quixey’s functional search to look for a specific song from Linkin Park, and it will serve you results of apps where you can hear the song, along with direct links to the song itself. Or you can search for a taxi or car service wherever you are, and Quixey’s results will not just give you a link of car apps, but order them in a list with the one closest to you at that very moment at the top.

Functional search will open for a private beta in one month on Android, with a wider launch planned for June. Other platforms will come after that, Quixey tells me. The private beta will give users results for food searches — specifically around restaurant listings. As Quixey continues to iron out the kinks in the service, it will add more categories in time for the wider rollout.

Of course, there are others that are also going after the same problem of discoverability within apps.

Everything.me is one example of a would-be competitor. Another is the mighty Google itself. The company launched its own App Indexing feature last year. Quixey is unfazed by the competition.

“We think it is great that Google has validated ‘open in app’ as a search feature, and that Google is promoting deep linking to apps,” a spokesperson told me. “However, it’s different because it doesn’t give users a deeper view into apps the way Google gives its own services, such as Google Places, and it doesn’t change the ranking – so that on mobile you’re still on page 4 if you’re fully SEOd.”

And what’s still to be seen is whether app makers want to play ball. Whether an app is based around advertising or a paid download, if you are giving people results right away, that could keep users away from downloading your apps altogether — thereby denying developers of the ability to get downloads or direct traffic. Questions like these will push Quixey faster to figure out ways of monetizing in formats that directly benefit those app makers.

But the fact that Quixey’s cracked the basic idea of searching across and into apps in a way that is not dependent on specific keywords, and sorts by closest relevance regardless of who publishes the app, is the kind of disruptive move that might force solutions to those issues by default.

How functional search works

The idea with functional search is twofold. First, Quixey, as led by founder and CEO Tomer Kagan, believes that apps are very much the future for how users interact with content today, but to date there has no effective way to search across them in the same way that search engines like Google or Bing have done for the Web. And yet this is what people are going to increasingly demand for apps as well.

Second, effectively Quixey’s search engine wants to become the platform for content consumption in itself, the link between a user needing information and then finding the app that will give it to them. Quixey calls this concept “search as a platform”, and it’s so committed to the concept that Kagan tells me Quixey’s filed for a trademark on the phrase.

But it wasn’t always this way.

“We actually had the idea to offer functional search three years ago, but when we took the idea to investors they thought we were crazy,” he told me in an interview yesterday.

So what’s changed? The app economy has exploded — there are millions of apps, both native and web, and they have become the go-to way that people are experiencing content on handsets. But in many regards they seem as siloed as ever, and so the time seems more ripe to tackle the issue. (And that’s something Quixey’s investors think now, too, judging by the recent $50 million round.)

Some interesting points about how Functional Search will work: it won’t be based on paid placements (not initially at least) — sponsored results is something that Quixey introduced in its more basic version last year.

“There are no sponsored results in the results,” Kagain tells me. “I won’t say there never will be, but right now we still need to find the right path.”

Nor will this be about special relationships with particular developers based on any other kinds of deals. “This is just how we think search should look,” Kagan explained.

And while Kagan says that quite a lot of search platforms today have more human involvement than you would have guessed, the idea with Quixey’s semantic crawling is that it is entirely based around an algorithm that will crawl deep links within apps themselves.

“Quixey’s Functional Search is based on just that, functions,” Kagan says. “We don’t help users just find apps, we are providing results based on their functionality as to why it is the best answer for them. Fundamentally, we will prioritize who gets to the top of search listings based on how well we believe the deep link answers the query. Additionally, ranking, usability, a bunch of variables also help to prioritize who gets to the top.”

Kagan says he and Quixey are fundamentally different in viewpoint from the likes of Google or even Siri. “Schmidt once said that the ideal Google result would be a single answer,” he notes. “And increasingly that single answer is one of Google’s. I don’t believe Google is in the real search business anymore.”

While there will not be any special deals with developers, what Quixey has been doing is talking to different publishers across categories like gaming, food, entertainment and so on to see what kind of results they think would work best for their own specific types of content. Links to music or specific recipes are fairly obvious, but how do you best represent gaming searches, for example? The answer there, for now, looks like it might be links to specific ‘rooms’ in a game that take you directly to playing. So in the case of poker, a particular tournament or list of active tables.

Allthecooks 3