Apple’s App Store at the end of the app era

Like a fly frozen in amber, the App Store’s fundamental deal has remained unmoving since its inception. As an enormous ecosystem swelled around it, crystalline structures of new rules and avenues of customer interaction have also grown — but not nearly fast enough for most developers. Especially the “long tail” of independent app makers which account for the most vociferous and loyal segment, if not the majority of downloads.

Then last week, Apple made three big announcements about changes to the App Store. It would bring search ads to the store, make review times much faster for developers and allow any category of app to use subscriptions — paying those developers an additional share of the revenue if people stayed subscribed more than a year.

Why those changes were announced before Apple’s Worldwide Developer Conference, which begins tomorrow, is a matter of who you ask. People inside the company with knowledge of the keynote tell me that it was simply too crowded to spend the time talking about these changes and the nuances that go along with them. People outside the company who are generally well-informed have suggested to me that Apple wasn’t sure how violently the developer community would react to changes that are controversial to some.

Whatever the reason, the past few days have given developers time to think about them, people time to complain about them on Twitter and journalists time to generate a hundred hot takes.

I have some thoughts and some additional details about the announcements now that I’ve spoken to people who know how they’re actually going to get executed, so I thought I’d put them down before we get swept up in the week ahead. I’ll just take them one at a time.

App Review Times

App reviews were always supposed to take five working days. They often ended up taking far longer, especially during the busy weeks or months before Apple ships new versions of iOS. It’s been a constant source of friction between Apple and the community. Now, the goal is to have 50 percent of apps approved in 24 hours and 90 percent approved in 48 hours.

Apple is so confident in these new times that it has removed the chart that told developers what review times to expect from its site entirely. I’ve been tracking the times and they seem in line with claims so far. An independent site that tracks times is also in line.

Apple has increased staff, changed policies (how strenuous it gets with known developers) and improved procedures and tools used to evaluate apps in order to make those times happen.

The changes to review times mean that developers can think about what kinds of features to roll out, bugs to fix and updates to make on a far different time scale. It could now take as little as a single day to roll out a bug fix or a major update to your app from the minute you had it locked internally. Though review times are getting a lot less play than other changes, I feel that this could have a significant impact on how often apps get updated and iterated on. As an independent developer, you may not be able to launch a major new version every month like Facebook, but you now have more options.

Subscriptions

From the beginning, developers who sell on the App Store get 70 percent of the revenue from an app, and Apple takes home 30 percent. Last week’s announcement cracked that amber for the first time, allowing a tiny fly foot to wriggle free.

Now, developers who institute a subscription fee will see their profits swell to 85 percent in the second year of that subscription. If they keep customers around, they’ll see more money. And to facilitate that, Apple also allows any app category to take advantage of subscriptions.

There are now 200 pricing tiers, though those won’t be revealed until later this year. The revenue split changes this week. Developers will now be able to allow users to upgrade, side-grade and cross-grade from one package to another very easily, and Apple is looking at ways to make the interface easier than it currently is. There is a notification when a subscription is going to be renewed and renewal is opt-in, rather than opt-out. You’ll have to say for sure it’s working for you.

The widening of subscriptions to all categories isn’t an exact replacement for a free-trial situation, but it offers a much better situation than before. Any app that decides to roll out a subscription can now offer its wares for free for a time until it requires a subscription, or to a subset of options. These already existed in some apps like news and music, but games developers — who occupy the biggest App Store category — will now be able to play with these tools.

The logic behind the subscription model and the split changing from 70/30 to 85/15 in the second year is simple: It incentivizes developers to continue to maintain and develop apps, which results in fewer dead-end apps; and it provides a boost in revenue as apps with back-end services scale up and retain users, allowing them to offset costs.

This is why I found Google’s somewhat suspiciously hasty announcement the day after Apple’s a bit less than fully considered. By increasing the cut from day one, you’re encouraging every app to be subscription-based, whether that’s actually the best model for that app. There are, and will continue to be, thousands upon thousands of utility apps that you need to buy once and use either regularly or occasionally. They don’t live on your home screen, but they’re there when you need them and they accomplish a simple task without any ongoing maintenance beyond updating for new versions of iOS.

MileIQ, an app Barnard points out already does well on subscriptions.

MileIQ, an app Barnard points out already does well on subscriptions.

Yes, from here on out, the biggest apps will likely all be subscription-based or free with other sources of revenue. But the long tail of interesting capsule apps, whether games, productivity, business-related or just utilities — those apps will disappear or  they will have to make up for revenue they would otherwise lose by being forced to adopt a subscription when it makes little to no sense.

I think a lot of developers (and other industry folks) who don’t follow the App Store super closely would be surprised at how many apps are already doing quite well with subscriptions, even ones that don’t have killer services and/or content,” says David Barnard of Contrast. “I think a lot of people are freaking out prematurely. Subscription fatigue has already set in for many with Apple Music, Spotify, Netflix, Hulu, etc. But that’s quite different from paying $3-$5 for an annual subscription to an app like Launch Center Pro. I doubt enough apps will go subscription for the average user to notice much difference. Maybe $10/year total. And the thing is, subscriptions necessitate some sort of free trial or free tier, so people will get to check out a few apps before subscribing to the one that works best for them. With paid apps, you end up spending $10 trying a few apps, and only ever use one.”

“I do think the onus is on Apple to help prime the pump for this transition,” continued Barnard. “Developers who choose subscriptions need to make sure their apps are truly delivering value over time and worthy of a subscription, but it would help substantially if Apple worked to normalize and validate the idea of subscriptions (and the value of software in general) in some sort of public facing way. Something like the Your Verse ads, but with some more direct hints at the value of apps to our lives.”

Search ads and discovery

Apple is making a couple of changes to the App Store’s discovery mechanisms, in addition to adding ads that appear in searches for apps. It’s bringing back the Categories tabs, for one. It’s also beefing up the Featured tab with better intelligence, filtering out apps you already have installed. Some iOS software betas, by the way, filtered the top charts, as well, but that is a bug, not an intended feature. Top charts will always show apps you have installed as well as those you don’t.

Here’s another fun one: If you 3D touch (hard press) on any app on your home screen, it will soon, by default, have an action that allows you to suggest that app to someone. This is a nice addition that will boost discovery via word of mouth and also provide a basic action anyone can take using 3D Touch, which many developers have yet to implement.

Now, about those ads.

There’s a public stat quoted by Schiller that 65 percent of downloads come from searches on the App Store. I’ve clarified, for what it’s worth, that this is 65 percent of all iOS downloads, and not some subset that only includes downloads originating on the store. That’s a pretty big target. And this is why Apple went after search ads first — it was the biggest possible target for revenue and for developer promotion. Screen Shot 2016-06-12 at 2.35.25 PM

Apple’s ad system is heavily based on relevance above all other factors. Relevance begins with an app’s metadata, which is used to automatically create the ad. The resources for the ad also come from an App Store’s pre-approved listing, which makes it tougher to game the system.

This relevance setting means that if you’re searching for a racing game, a word search game will never ‘win’ in the ad auction process for that term. Apple is also attempting not to penalize new apps in categories for not having a lot of ratings, giving them a chance to win ad auctions and appear in categories. And all apps, of course, pay only on tap not on sight. Apple is exploring things like re-engagement ads for the store as well — so you could be able to direct people to a specific point in your app even if they already use it.

The relevance metric, though, is the one to pay attention to. Specifically, something Apple calls the “minimum relevance threshold.” If your app does not pass this threshold for the search that you’re trying, the app will not even get into the auction. Apple will average out the relevance scores across ads for that term, and if your app doesn’t pass a certain mark of relevance based on keywords, reviews, downloads and other factors, it won’t be allowed to clutter up “non relevant” categories.

A few privacy notes:

  • Apple is not creating search profiles (a very common practice with most ad products) for users, but they will be looking at data internally on an aggregate level to make sure they’re showing people relevant ads. That includes things like whether you’re clicking on ads, what you download, your age, your location and your gender.
  • Apple will not be merging its own data with external sources like Axiom or others that collect user data across networks. And Apple does not share that data with external networks or other developers. You won’t search for an app on Apple’s store and then see it in a Google ad.
  • Tapping on “Ad” on any App Store search ad will also allow you to see the exact data that was used to show you the ad. Which is very nice for transparency.

Apple’s existing ad products like iAd, which didn’t work out so great, will still contribute technology, data and findings to this new effort. So even though it failed the first time, Apple still gets to use the tech to build this out.

Apple’s search ads program is available for free during beta. So if you’re an App Store developer, you should do yourself a favor and try it out during this period, which starts Monday. The ads may be tests (that Apple is watching to prevent gaming), but the downloads are real. The program launches for real in the fall.

Apple is doing its best to minimize the complexity for indie developers. Apple’s behind-the-scenes tool will build your creative automatically, select your keywords and set up the auction for you. All it asks for is a credit card and budget. The big players have more tools available to them, but this ease of use means that any developer should be able to try the system, regardless of how much ad experience they have.

No developer, no matter how big, will never be able to buy exclusivity or a ’blackout’ in a category or search. And there are no minimum bids in the Dutch auction system.

Developers I spoke to seemed optimistic about whether Apple’s search ads would move the needle for them.

Search Ads in the App Store allows us to reach users exactly at the point when they are looking for apps. It is a relevant environment and the App Store is truly at scale, which very few advertising channels are, especially for mobile,” said Marcus Gners of Lifesum.

Mathieu Nouzareth of FreshPlanet which makes the music trivia game SongPop, shares the sentiment.

“Every app developer is confronted with the same issue: how do I make my app/game known to my potential users once I launched it?  I have seem estimates/reports showing that there are about 1,500-2,000 new apps being launched every single day in the app store! Discoverability is an issue for all game/app developers and there are only a few ways you can have you app being discovered:

    a) You pray and hope that a miracle happens

    b) You are featured by Apple

    c) You reach out to journalists

    d) You take matters in your own hands and you decide to invest in marketing campaigns. 

Until Facebook mobile ads came along, it was not very practical for anyone to do efficient mobile install marketing at scale. No one was offering the holy grail of mobile installs marketing: the combination of reach, targeting and reporting. Based on what I read, I think Search Ads will be able to combine those 3.”

That was a common theme when talking with developers, by the way, and I think that Facebook will less be threatened by this than they will be bolstered — making developers more familiar with search ads will only increase their curiosity in other products of its type.

“In terms of cost, it is going to be interesting to see the actual price on Search Ads versus other solutions,” Nouzareth continues. “Apple claims that the their solution will be cost efficient because ad relevance is prioritized over spend. This would be very appealing for smaller app developers as CPI (cost per installs) tend to rise dramatically on other platforms, thus favoring bigger developers and those who have apps/games with the highest LTV [life time value].”

Neal Shenoy of Speakaboos says that search ads may help developers because they’re based directly on download intent. “…Consumers are being more selective and downloading less apps in general increasing the cost of app marketing,” says Shenoy. “The result is that discovery in the App Store is a true “last mile” problem that we believe search ads can help solve. Whether an app developer is looking for new customers or re-marketing to existing prospects, Search Ads should accelerate download volumes, generate more qualified prospects and improve conversion because they are being delivered at the precise point of consumer intent.”

Changes

The current temperament among developers is that if this is “it” for the App Store in 2016, then it is not enough. Many indies, especially, are hoping to see a re-energized Apple now that control over the App Store has moved under SVP Phil Schiller and away from Eddy Cue.

Cue, Apple’s content deal-maker and broker of the Beats deal, has become a hugely polarizing figure among developers. Many feel that Cue was not giving the appropriate amount of attention to pushing the store forward, focusing instead on deals with content makers, celebrities and other power players.

There is a cautious optimism among the community that Schiller will right the ship and shatter the stasis that the App Store has endured over the past few years.

“Overall, it’s not like I think there’s some magic bullet that will make everything okay for indies, so I’m glad the person in charge is thinking about us and trying a bunch of things,” says Phill Ryu of Impending.

Everything I’ve been hearing is that this is the “first of many” changes to the App Store. Some of those changes will come this year and some will come later, because there are many structural challenges in rolling out major changes to a store that serves hundreds of millions.

In my mind, the changes are linked. Subscriptions add much more flexibility across the board to enormous categories of apps that cannot be defined specifically as “subscriptions to services.” Improvements to discoverability (as long as they continue) aid in the long tail still being a place where business are built. And shortening review times creates a faster moving and more iterative App Store mentality.

But I think that’s just the beginning. And Apple needs to acknowledge that these improvements are, frankly, just table stakes in comparison to many of the things that Google has announced recently to developers on the Play Store. There is a new era on the horizon: the era of the distributed app.

I’ve written before about “invisible apps,” which offer benefits to the user without being on the home screen, being opened or even having a “home” on the phone. Recent years have shown that this is the way things are going. People download fewer apps and keep the ones they use more. But they use more services than ever.

We’re headed for a point at which people use more “invisible,” distributed services on iPhones than they do “apps you launch via an icon.” That’s a given. So Apple has two choices — either it can milk the last drops out of the capsule app ecosystem as we know it, or it can start building App Store tools to support these kinds of apps and services and to help developers monetize them.

Subscriptions and services, of course, go together hand in hand. What if a developer could craft software that provided a truly useful service but had no icon, required no installation and was paid for out of an all-in-one App Store subscription that was based on usage? What happens when virtually and augmented reality are ubiquitous and phones are small again and the home screen means nothing?

That’s going to happen. It is happening already with bots. Whether Apple is flexible and attentive enough with the App Store to allow it to capitalize on this new era is the only question.

So, these changes are baby steps, but they’re baby steps with the right fundamental mindset. Developers should hope that the trend continues.