What Facebook’s AppLinks and Unbundling Can Tell Us About the Mobile Web Ecosystem
I heard a great joke the other day by comedian Ronny Chieng about apples and oranges (first 44 seconds; the rest is about Kanye West with NSFW language).
It turns out you can compare apples to oranges. I was reminded of this joke after last week’s two major mobile developments. First, Foursquare announced it was unbundling its app into two—one into a Yelp-like clone and the other into Swarm, a location-based heat map. Secondly, Facebook introduced AppLinks, the open source project to allow deep-linking between apps. Both developments are indicative of how much the mobile ecosystem is still evolving and how it’s evolving differently from where the web was 20 years ago. These two platforms are as different as apples and oranges. But as Ronny masterfully did in his bit, I’ll try to compare the two and hopefully, provide some insight on where I think the mobile ecosystem is going.
Nothing Beats the Native Experience
Consumers overwhelmingly prefer native over the mobile web. A recent study from Flurry showed that native apps trump the mobile web 6 to 1.
Press enter or click to view image in full size
There are a couple of well-documented reasons for this. Native apps tend to be much faster. Native apps also can utilize the features of your mobile phone including the camera, GPS, accelerometer, push notification and address book. For these reasons, the user experience is typically superior. Apps also provide better monetization. A recent study by Flurry and IDC showed in-app revenue outstripping PC web revenue by 2017 and growing faster than their mobile web counterparts. All this may explain why consumers preferred the Facebook app experience over the mobile HTML5 version. But why did Facebook choose to separate Messenger out of the core app? One reason is by unbundling, both apps can live on the home screen and be accessed with one tap instead of two. In my opinion, there is another user experience reason but in a somewhat duh sort-of-way. It’s because the screens are small.
An iPhone 5 display’s dimensions are roughly 3.5 inches by 2 inches. Compared to my MacBook Air display and ThunderBolt display, it’s tiny. My thumb, on the other hand, is 2x larger than the mouse cursor (it’s not a fair comparison, like apples and oranges). Here are what the various screens and touch points look like next to each other.
Press enter or click to view image in full size
Suffice to say, the mobile device’s small form factor is compounded by its larger touch point. This means when you have a button or icon, it has to be small enough not to overwhelm the rest of the app, but large enough to accommodate your thumb. All the mobile OS makers have suggestions on how big to make this minimum touch target. Taking that into account, having an additional feature like messaging in your core Facebook app can yield a poor overall experience. Consider the original Facebook app with messenger in-app. It used to have a pop-up show up with your friend’s face on it. That pop-up took up about 5% of the whole screen. The messages link on my Facebook web page takes up about 0.5% of my screen on my MacBook Air and 0.09% of my ThunderBolt display.
On mobile, all the apps are in full screen.
Even with mobile multitasking, the experience is not the same. On our laptop and desktop, apps are in windows or tabs and navigation between them is easy and quick. This is why Facebook, Google, Dropbox and arguably now Foursquare are splitting their apps into single core functions. Adding one additional function may seem harmless on a web app. On mobile, though, it can literally take up 5-10% of the core app’s visual experience, distracting the user from the core function. Consequently, it devalues the user experience of the secondary function as well.
If we’re unbundling functions into discrete apps, the App Store experience becomes even more meaningful.
The App Store Keeps the Web Away
The App Store is revolutionary. While the internet struggled in its early infancy on monetization, for mobile app makers, monetization comes easily through one-click purchasing. So far, the majority of mobile engagement is from gaming (32%) and social (17% for Facebook, Social Messaging at 9.5%) but the opportunity for news and productivity is palpable as more young people bring their mobile-habits into adulthood.
The App Store also has its drawbacks. There’s a bit of a mid-90s Yahoo! feel to our App Store experience today, with its categories and recommendations. Paying special attention to the hierarchal nature of both landing pages, take a look at the 1995 Yahoo! home page and today’s App Store. Today’s App Store is a bright and shinier version of the Yahoo! home page, catered to everyone and no one in particular at the same time.
Since most of our time consuming mobile content is through apps, the App Store almost acts as the true mobile browser, with its home page permanently defaulted to Apple’s landing page. In a way, Apple’s App Store experience is not too dissimilar from the historical Microsoft web experience: A consumer would power on a Windows PC, launch Internet Explorer and see the MSN home page for content.
In the App Store, Apple’s gatekeepers hold the keys to app discovery by curating and recommending their favorite apps. In a recent article on Slate, Jim Edwards observed that at most, only 200 out of the 1 million-plus apps on the App Store are shown in its display (or 0.02%) at any given time. If that seems somewhat arbitrary like Yahoo!, then maybe search is the answer.
As we all know from personal experience, app search is comparatively poor to its web counterparts. For example, a search for the Medium app on my mobile phone jumped right to the app but it was non-existent on my iPad because Medium is iPhone-only. I didn’t know that but because I’m a more sophisticated App Store user, I figured I had to search for it under iPhone only. I suspect many iPad users are not as sophisticated. In an ideal world, the search should take me to the preferred app version depending on the device I was using and only choose the less desirable version if the first choice was unavailable.
Once upon a time, search on the web was just as difficult. Remember when we had to string boolean search terms into Google just to get the answer we wanted? This opened up a cottage industry of search engine optimization experts (now called inbound marketers) and analytics tools. We are already seeing this on mobile with vendors like App Annie for app store optimization and analytics.
Today’s App Store experience feels like 1995-2004, post Netscape but before the sophistication of search. The issue of poor app discovery is compounded by the recent trend to unbundle apps, which should flood the already under-water App Store with its million-plus apps. What we need is a modern mobile browser that can seamlessly launch a native app experience alongside mobile web content and with a powerful search engine behind it. Could AppLinks help bring this about?
The Browser Won’t Be Second Banana
AppLinks or Deep Linking is not particularly new. Facebook has already been doing it within its own app ecosystem. Android apps already had deep linking enabled, including for search results. A few startups like URX are utilizing deep linking within apps for highly targeted mobile commerce. The move towards standardizing deep links with the free and open-source App Links means cross platform, web to mobile and app to app linking could finally be possible.
The first obvious case for deep linking is in mobile ads and commerce. Mobile ads with deep links are already being used at Facebook, for example, in their mobile app engagement ads. URX is building a business around this concept. Search is also empowered by deep linking. Searching with Google or app search engine Quixey (who developed their own standard for deep links in 2013) will already be a dramatic improvement over the search capabilities in the App Store. It also solves the problem of finding apps who don’t make it to the homepage of your OS and are deeply nested in groups. But all these are baby steps towards what I think will be App Store emancipation through a new mobile browser.
Right now, app makers are beholden to the App Store (and by extension, Google Play) because they hold distribution, one-click purchase and generally all the UX advantages you would have being a native app. This means app makers will continue paying the App Store 30% tax. But what if there was a mobile browser designed with deep links in mind and powered by a sophisticated search engine to find these apps? This browser would need a user interface that doesn’t distract from the app’s content and can seamlessly move from web to app and app to app. Since app linking is cross platform, the browser can operate on any OS, potentially solving the distribution advantages of the App Store. Deep links may also enable redirects to payment processors outside the App Store’s jurisdiction, freeing developers from the 30% fee (Note: this is where it gets legally tricky. Ask your lawyer if breaking TOS is right for you.).
While it’s unclear if the promise of deep linking isn’t more than better UX, there’s a lot of unknown upside in this cross platform protocol. It may even make HTML5 apps more viable.
App Links is not the panacea but it could be an important first step towards unlocking the gates around Apple walled gardens.
The mobile app ecosystem hasn’t and won’t develop like the web. So far, it’s done some things backwards like solving monetization first, which is a good thing. In our rush to market though, we’ve ceded initial control to gate keepers in the form of fees and censorship. But history has shown that gate keepers can’t keep technology and innovation stymied forever—whether it’s in the software, music or movie business. I understand these analogies aren’t perfect and I’ve simplified some of the arguments here. But whether I’m looking at apples to oranges or Apple to Google, as Ronny has shown us, it’s not impossible to draw comparisons and form our own conclusions.
Notes:
- Special thanks to Jeff Chen, Rich Yueh, Jon Thornton, Duncan Carroll for reading and commenting on this post.
- I used Apple as the proxy for my argument. Google’s garden is less walled, but they still own much of app distribution which is relevant. Again, apples and oranges folks!
- Images from here, here and here.
- If you liked this, follow me on Twitter: https://twitter.com/DavidPCheng. I tweet stuff on tech, startups, inbound marketing, growth hacking and Liverpool FC. My Medium blog posts are on startups and tech (which are linked below).