Deeplinking -- And 4 Ways To Fix App Discovery
I’ve been spending a lot of time over the past couple of months thinking through why there’s no great way to discover mobile applications in 2014. There’s been a number of posts on the topic, including why we’re still in the pre-pagerank phase of mobile. Facebook also brought the issue more into the spotlight with AppLinks yesterday. The problem I see is that we’re trying to correlate something from the desktop world — web search — to the mobile world — app discovery. It’s a wildly different problem and here’s why:
Verbs vs. Nouns+Adjectives
Discovery in the desktop world was primarily driven by searches for nouns and adjectives. i.e. — best place for asian food as opposed to what we’re seeing in a mobile world, which is primarily driven by searches for verbs/actions i.e. — track my weight. Mobile apps are primarily utility based pieces of software as opposed to content driven experiences that dominated the web. We access and utilize them based upon verbiage — bring me a private car, track my weight, order me food, book me a hotel tonight, and much more. We aren’t looking for singular pieces of content like we did with web search, we’re looking for verb driven software. Discovery of apps is driven by a user wanting to leverage a piece of software to utilize a verb, not find more nouns/adjectives.
People vs. Algorithms
In a desktop world, there were indeed utility based pieces of software, yet they weren’t really discovered via “search.” Think of the valuable pieces of software that started in the desktop world: Facebook, Salesforce, Twitter, 37 Signals, Mailchimp, or any other web app. There were no algorithms to discover the latest and greatest web apps. We discovered software in the desktop world primarily through other people. We’d hear from the press or we’d hear from our friends what software works great. Very few people have signed up for Facebook or a business piece of software like Salesforce by Googling for it, instead they hear about it from a friend or a trusted source. Discovery of apps will likely be people driven over algorithms.
Deeplinks are for connecting actions, not indexing content
To take a “Google GOOG +0.05% like” approach to solving app discovery we’d have to assume that there was enough indexable content that matters to a user. Yes, ECommerce apps like Gilt have content and social media driven apps like Pinterest have indexed content. We already know about those companies and they were not created mobile first. Their content is already indexed anyway through the mobile web. The mobile first companies like Uber and Hotel Tonight have very minimal indexable content. Yes you might be able to index the driver’s profile page or the hotel’s listing on Hotel Tonight. That’s not what people looking for the app want. What they want is the ability to make a verb happen. Deeplinks are to verbs, what hyperlinks were to nouns/adjectives. Don’t confuse deeplinks with having the same purpose as “hyperlinks” on the web and indexing. Deeplinks are going to let apps pass actions from one to another. I save (verb) to a venue on foursquare and then discover I can book (verb) a meal on open table, where the deeplink opens the opentable app to do so. Deeplinks are going to string together verbs, sort of like a giant IFTTT chain. Users are often looking for a piece of software that allows them to take action, as opposed to a piece of content.
Millions of apps in a mobile world as opposed to thousands in a PC world
We are seeing a software explosion like none other. Yes, the software industry has been around for decades starting with the Desktop PC era then the Web app era. The mobile era is orders of magnitudes larger. Anyone can create a piece of software, which means we’re now dealing with millions of pieces of software. We never had to solve “software discovery”, at least not on a scale like this. Web search solved content discovery, app search needs to solve software discovery.
I don’t know what the solution is here, but I know it’s a problem that needs to be solved. There will only be more software as time goes on, especially as more mobile devices get sold and more capabilities are built in for developers. To sum up the points above, I’d say the solution is likely to be built on the following pillars:
- Focusing on verbs and actions as opposed to focused on adjectives/nouns
- Powered by people and trusted sources as opposed to purely algorithms, though those will be baked in.
- Focused less on indexing content and more focused on linking together verbs via deeplinks.
- Focused on discovering software, not discovering content.
What are the best examples of startups attempting to tackle this or what ideas do you have?