#app linking

Last week, I was on my way to meet a friend, and decided to go a bit early and get a haircut first. as I was in desperate need. It was a new barber that I had not visited before. I got my cut, and toward the end, I realized I had some extra time to kill. So I asked my affable new barber if there was a place to get coffee nearby. He gave me short directions to a nice little cafe nearby. I was able to get a caffeine jolt and answer some emails, and the whole experience was really nice! I killed two birds with one stone, and was ready to go on and track my friend down.

Our physical-world interactions offer a ton of opportunities for product extensions like that. If you go to a store to do some shopping, you can be directed to other stores nearby. When you get your car serviced, you can be pointed to a place to get a bite to eat while you wait. And you can buy a toy for your kids, and you’ll likely find batteries near the register.

The world of native apps is vastly different. Early apps were prototyped and built in an atmosphere that was largely void of a need for cooperation or interconnectivity (the fruits of this vestigial limitation are seen when even the most tech savvy among us are all totally blown away with simple links between apps). Additionally, thanks to the proliferation of the on-demand economy, many apps are becoming increasingly sole purposed uni-taskers.

Translation: we each have a ton of apps, but the functionality of each is generally very limited.

Kinetic Intent

OK, so what does this mean for app users? Well let’s look at a pretty well-distributed use case like a food delivery app. Typically, I use Seamless (my go-to) to order dinner to my apartment. Once I add meals to my cart and confirm the order, I have now completed 100% of my need for the Seamless app at that moment. I’m not likely to tap back into the app content to order more food (taking a page from Chris Maddern’s ever-insightful points on this subject).

So what is Seamless to do to make a better experience? 

Currently, there’s little call-to-action to do anything else once you’re done with your order. So, Seamless is functionally telling me to just close their app and do something (anything) else outside of their universe.

But that approach leaves me with a kinetic intent, an opportunity to engage with other apps, but no natural path to get there.

Constellations and Intent Indicators

The Deeplink Team has been banging on about app constellations for years. 

Long story short, the apps you have on your phone are a huge indication of what you may want to do. We each curate our own unique fingerprint of apps that we see as the reputable authority figure on a given subject (travel, taxi rides, news, restaurant reviews, movie tickets, email, reference, music, reading, maps, and every other subject). Each member of our personalized network of apps says a lot about what we may like to do. I may use TripAdvisor as my travel reviews app, while my friend may use Yelp. So a single link to either of these properties just wouldn’t make contextual sense.

What would make sense? All of these links should be personalized and dynamic. Each of us should be linking to entities within an app that we are most likely to interact with!

Linking Out

Taking the above as a given, there’s still the underlying issue of linking out for an app. This idea of sending users out of your app is still incredibly nascent.

App usage is sacred, so developers and marketers are understandably precious about assisting their users in leaving their carefully crafted native experience. To most marketers, it seems counter-intuitive and counter-productive (“I live and die by DAUs, MAUs, and time-in… Why would I ever send a user away from my app?”).

But the answer is amazingly simple. Your users are leaving anyway! Not only are they leaving, but even worse, they are leaving without direction, and you’re getting nothing for that valuable pass-off.

About a decade ago, Yaron Galai founded what has become a hugely successful content discovery company called Outbrain, preaching a new and previously foreign mantra to web publishers:  

"If you love them, let them go."

Take my barber example from the beginning of this post. If when I asked where I could get a cup of coffee, the barber had just stared blankly at me without providing a suggestion, or said “you’ve already spent money with us, so get the %*@# out!,” it’s safe to assume that I wouldn't likely return to that barber.

Why are apps any different? They shouldn’t be. Developers should help users find the next app they may want to engage with, harnessing the intent that is complementary to the action that they most recently completed. Soon enough, users will start forming habits around those connections. In my use case, perhaps Seamless, on their order confirmation page, could suggest Netflix movies I may want to watch during my evening at home. That may make me more likely to use the Seamless app to access that utility more smoothly in the future. Not only will developers be significantly improving their UX and embedding themselves in habitual behavior, but perhaps that next app will be willing to pay for that engaged user!   (hint to a future business model)

Exit Points

Once we accept that our users will eventually leave the app anyway, and there may be upside in helping them through this process (both monetary and otherwise), the next question that would occur to a logical developer is “where should I put these outbound links?” Specifically, what’s the natural exit point for an app?

The answer is quite clear for commerce apps like Seamless, Uber, OpenTable, or Walmart. Of course, once the purchase is made, the intent has dried up and there is in most cases limited need to attempt to keep the user inside the app.

Things get murky though when it comes to content apps like TripAdvisor, news apps like CNN or BBC, or a social media app link FourSquare. There’s no obvious point of universal intent exhaustion, and even the metrics comparison may be different! Some users may spend 5 minutes in the app while waiting for the bus, and others may stay for 100 page views. Some users connect sporadically, and others are incredibly engaged and view content frequently throughout the day and night. Some may be easier to re-engage with push notifications, email marketing, and advertising, which may make a developer more comfortable with the idea of sending them away. There’s no one-size-fits-all here.

We’re early in this experiment of app-to-app links, and every developer and/or marketer will need to figure out their own formula. But one of the biggest benefits of native apps is that the engagement is incredibly personal and customizable, and user behavior is so trackable. This makes it very easy to do some A/B test to find what works for your user pool.

The early adopters are going to have a massive advantage, as they are going to quickly arm themselves with a well-honed strategy for driving revenue, engagement, and usage (another key note, the more apps that drive out, the more chances all of us will have to drive back in).