Deep-linking technology has been making its way progressively into the mobile ecosystem, providing an easy way to take users from one app to a specific point in another one, much as hyperlinks do between web pages. Now one of the earlier movers in this area, a startup appropriately called Deeplink.me, is adding a new feature that paves the way for deep linking, and automatic searching, within apps themselves.
Today the company is taking the wraps off an SDK for iOS apps called the AppWords Assistant, an in-app search feature that will let developers call up content within their apps for users based on the actions that they have taken in the app in the past. The Assistant will be free to try out for the next few months, with pricing tiers based on number of clicks and deeper analytics getting introduced later.
The premise with AppWords Assistant is that even with predictive texting, voice and other features that make inputting information more frictionless for mobile users, it’s still a hurdle to get the majority of users to make a lot of proactive actions on mobile when they are in search of something, or when they want to perform an action on a phone.
The AppWords Assistant is built to bypass some of that by using data about your past behavior in the app, as well as in other apps, to figure out what you might want to do next. As co-founder Itamar Weisbrod describes it, the feature is somewhat akin to Google Now on Tap, the Android feature that predicted what you might like to look at based on what you are already doing on your phone (now being replaced by Google Assistant).
The new SDK is a variation on the company’s earlier AppWords launch, which aimed to build a bid-based ad network (AppWords is a play on Google’s AdWords) around deep links within apps.
AppWords’ linking between different apps didn’t really gain as much traction as the startup had hoped because of what Weisbrod put down to the challenges of building a two-sided marketplace.
It turns out that most developers are more interested in keeping users in their own apps, not sending them to other apps, although they were very interested in building links to try to bring more users to their apps. And so the earlier linked-out part of AppWords has now temporarily shut down. It is set for a relaunch later this year, after what the startup hopes will be a stronger take-up of the in-app search service that is launching today to lay more groundwork.
All the same, Deeplink.me has been quietly growing and generating revenue from its business, which now works with “tens of thousands” of apps using its code for deep linking to specific places in apps, with customers including Draft Kings, Urban Outfitters, NBC Sports and Yelp.
Many of the use cases today are in the area of commerce, where apps use deep-linked cards to prompt actions related to transactions. These app have the most need to convert users to both stay on their sites longer and hopefully buy things. “Transactions have the easiest ROI,” noted Weisbrod, since many commerce apps currently lack the mechanics to send users to relevant content. For example, a travel app where you have booked a flight should know whether you might want to book a car, or a hotel, or a restaurant, next, and provide the right prompt for the right person.
Deeplink’s growth as a relatively bootstrapped operation — since opening for business in 2010, the company has raised only $3 million in funding from investors that include Yossi Vardi, Flint Capital and Radiant Venture Capital — stands in pretty stark contrast with some of the other startups out there that have built (or tried and failed to build) businesses based around deep-linking technology.
Some of these have fizzled: Quixey, which had raised more than $134 million from investors like Alibaba, SoftBank and more, in February said it was exploring strategic options and now is rumored to be shutting down altogether. URX — another hopeful with a long list of well-known investors behind it — saw its team get hired by Pinterest, but the social networking platform did not buy the startup’s technology.
And some are still in play: We’ve heard that Branch Metrics, which has already raised more than $53 million, is currently looking to raise more than $50 million more; and Button closed a $20 million round this past January.
Weisbrod would not comment on where his own startup is standing with regards to its next funding round, although it’s safe to assume that it will likely try to raise more to both build out its platform and for business development.