The future of app stores

App stores, pioneered by Apple, aim to provide ease of payments and discoverability to developers and consumers. However,

  1. do they offer unique advantages that no other payment or discoverability system can match?
  2. is Apple’s success transferable to other app stores?

The answer to these questions is generally assumed to be Yes, but I’ve instinctively disagreed with that optimistic assessment for a while now. I just couldn’t quite explain why. So I was very glad when I found a clearer articulation of my doubts in a piece by Fred Wilson, a venture capitalist who has seen the light on HTML5 apps. He says:

Many of those who disagreed with my conclusions [that developing solely for iOS is not the right strategy] focused on what is working in iOS today. And my counter to them is what is going on today will not be going on a year from now and certainly not five years from now.

Sure, the system works. Now. For Apple. But what does that say about next year? Or about other platforms?

Clarification: From the reactions it appears I was unclear about one point: this article does not discuss native apps vs. HTML5 apps. Native apps can be distributed via the web; HTML5 apps via app stores. It’s the distribution medium that is the subject of this article; both native and HTML5 apps that are distributed via app stores will run into the same kind of problems.

Payments

App stores provide a useful way of making mobile payments. However, as we saw in the past few weeks, consumers do not pay enough money to allow a significant amount of developers to earn a living by creating apps. Besides, consumers want operator billing anyway.

So from a payment perspective the app store offers nothing unique.

Distribution

As a distribution channel, app stores are inferior to the web. The web already is everywhere, and app stores offer no unique value that trumps the web’s ubiquity.

In fact, app stores are severely counterproductive when you want to update your app: either you wait for ages for a murky process to end, or you have no control over whether or not to charge money for your update, or both.

Discoverability

But the point of app-store distribution, it is said, is your app being discovered by users. This might have been true once upon a time, but it’s rapidly becoming a rubbish argument — especially for Apple. There are just too many apps.

148apps found that currently the Apple app store receives 978 apps per day. Nearly a thousand! How is your app going to distinguish itself in that vast sea of stuff?

There are, in fact, several answers to this question:

The first two options cost you a lot of money. The last one doesn’t necessarily, but it’s so very hard to pull off — especially the bit about creating an app that’s like no other apps. With a thousand apps per day, few are likely to be unique — or stay unique.

Besides, a traditional market campaign would also work fine via the web, as would disseminating a unique app. Arguably, discoverability would work even better on the web because the parameters for succesful web SEO are known, and those for the 100 or so app stores aren’t.

So the only unique discoverability feature of the app stores is the Top Ten: a paid portal page. Count me unimpressed.

Cost of ownership

One underreported feature of the app store model is the cost of ownership. App stores are actually costing their owners money instead of bringing it in:

It would be fascinating to put an exact number on the amount spent by the mobile industry to support proprietary third party application development in the last 2 years alone. The ever increasing complexity and number of these developer programmes is driving billions of dollars of spending each year, across events, staffing costs, administering the app stores and partnership marketing. Some, such as Microsoft, have even started to offer direct financial incentives to developers to create third party apps for their platforms.

It raises a pertinent question: how long can so many individual companies support so many individual app stores and so many individual platforms? The economics simply do not add up. Every company, with the possible exception of Apple, is operating their app store and developer programme at a substantial loss. This cannot continue indefinitely.

Sure, Apple makes a profit. So does GetJar. Google might. Nokia might in a while. But all the rest? LG and T-Mobile and Dell and Microsoft and Samsung and AT&T and Intel and ... well, look at the list yourself.

Eventually costs will be cut, and app stores will be closed. It’s only a matter of time. And if that happens, where will the companies turn to spread their apps?

The present of app stores

So here’s where app stores stand nowadays: They offer a payment system that works, but that excludes the majority of the world population. What they don’t offer is enough money to make a living out of creating cool apps. As a distribution system they handle app updates really badly, and the only unique discoverability feature they offer is a portal page. Finally, they cost a lot of money.

However, nobody is interested in this reality. Instead, everybody copies Apple because everybody else does. (That’s the definition of a hype, by the way.)

  1. Become app publisher (like Apple did)
  2. ???
  3. Make shitloads of money (like Apple does)

Let’s take a closer look at what they’re copying.

The iOS architecture enforces a distribution monopoly for Apple’s app store. This monopoly is used for controlling money, content, advertising, and discoverability, with the predictable consequence that Apple makes a succulent amount of money.

This business model reminds me of the publishing industry, and we all know what the web is doing to publishing.

Now don’t get me wrong. I’m not saying that Apple is making a mistake here. Anyone who follows the mobile market even a little bit knows that this model works fine for Apple.

Apple cleverly used its strong developer support and its loyal, affluent consumer fan base to create the most desirable mobile platform in the world around its publishing business model. And then locked it down tightly.

Copying Apple

What I’m doubting is whether others can copy Apple’s business model. Only OS vendors can lock down the system to create an effective distribution monopoly, and no other OS vendor has the position that Apple has with both consumers and developers.

Google attracts developers, but not consumers, and it had to hitch a ride on established mobile brands to reach them. That worked to the extent that Android is considered iOS’s most serious competitor, but it has made Google dependent on the mobile brands. Thus it cannot go too far in locking down its system.

Nokia, Samsung, and RIM attract consumers, but not developers, and I don’t see any solution for that other than patient outreach. If developers would flock to one of them they’d be in the same position as Apple, but I don’t see this happening any time soon.

The others do not own an OS or have a negligable market share. Well, Microsoft could try something on (desktop) Windows, but that would not be received gladly by device vendors and system administrators.

In other words: no market party has exactly the same position as Apple, and as a consequence copying the app store distribution monopoly does not make sense for them. They recognise that fact: nobody has locked down its OS as much as Apple has. But without such a lockdown app stores don’t really have a function.

The future of app stores

So most of the app stores are a solution without a problem, and many will be closed down because they’re too expensive to maintain. Oracle’s Java App store has already fallen.

However, the platforms that lose app stores still need a way of finding, paying for, and installing apps. Obviously they’ll turn to the web. Around the same time, it will be discovered that HTML5 apps offer a way of reaching beyond the narrowly-defined borders of one single platform. Fred Wilson, again, stated the problem most succinctly:

I'm not really focused on who makes more money in smartphones. But I care a lot about where our portfolio companies should be focusing their precious mobile development resources.

The web will win easily. If you want a success metric, keep an eye on the daily number of submissions to the Apple app store. As soon as it drops, you’ll know that the victory march of the web has started.

That doesn’t mean that app stores will disappear entirely. There will always remain a demand for them.

Discoverability-driven app stores will spring into being for niche markets with a relatively low number of apps — say, mechanical engineering. If it’s possible to scroll through the entire catalogue in five minutes, the app store has a unique discoverability value and will become a focal point for people that need these apps. It might even be able to survive financially.

Apple becomes unique

More importantly, the Apple app store will continue to exist. Its success is based on developer and consumer loyalty that isn’t going to go away any time soon.

As downloadable HTML5 apps grow in popularity Apple will eventually have to open the gates or risk becoming marginal beyond its fanboi core. As far as I can see Apple is aware of that — it created the best mobile browser in the world exactly to be prepared for this eventuality.

But even if HTML5 apps become ubiquitous native apps won’t disappear. Due to the peculiar nature of the iOS ecosystem, that means the app store won’t disappear either. It will become smaller, though. Apps might be discovered again!

The best part is that Apple’s app store will finally be seen for what it is: a unique ecosystem that cannot be copied by anyone else. Because it cannot be copied it has zero predictive value for other ecosystems, so it’s pointless to study it if you want to understand the mobile market as a whole.

Once that’s clear to everybody the rest of the mobile market will be free to figure out how the real app system is going to work. One thing is certain: it will be web-based.

This is the blog of Peter-Paul Koch, mobile platform strategist, consultant, and trainer. You can also follow him on Twitter.
Atom RSS

I’m speaking at the following conferences:

(Data from Lanyrd)

Categories:

Monthlies: