War of the platforms: Facebook, Apple, Android, Twitter.

For the first time in decades, the choice of what platform to build for is not obvious.

Back in the 80s and 90s, it was obvious: Build on Microsoft. Then from 2000 to 2008, the closest thing to a platform was Google, where developers would work with SEO and SEM tactics to get traffic. Then all of a sudden, the Facebook platform got big- really big. Then came mobile.

The last time this happened was in early 1980s
All of a sudden, you can actually pick and choose what platform to actually build upon. Weird. This is a historic event – the last time there were this many choices, we were choosing between Windows, OS/2, or the original Mac.

For those with deep pockets, of course you can build on all of them – yet if you’re an early startup, you really have to double down on one and go multi-platform as you pick up traction.

Evaluating platforms
To evalute which platform is best, here are some thoughts:

Which offers access to the most relevant users?
Which one is the most stable?
Which platform is most unlikely to build a competing app and try to replace yours?

Apple
Ultimately, I think distribution is where platforms really help. As Apple’s demonstrated, you can make developers learn a whole new programming language, a new technology stack, if you can give them access to millions of users. Contrast that to many generates of Google and Yahoo APIs which allowed for data access, but not distribution – much less useful. The biggest problem with Apple is that their leaderboard system is rapidly filling up with winners and it’s harder to break in.

Facebook
Facebook is much more of a free-for-all, and new apps can break in, but they are pretty unstable and are constantly changing their platform. The plus side is that their constant changes introduce new windows of opportunity for an adventurous developer to jump in.

Twitter
Twitter as a consumer product is so simple, there aren’t many marketing channels to even take advantage of. They don’t have an app store, they don’t have an apps page, and it’s hard to discover. Right now, as a platform Twitter’s not that great.

Android
Android seems like a potentially great platform to develop for, but there’s so much opportunity in the iOS world that most developers have overlooked it. Perhaps it’ll turn into the contrarian bet and we’ll see some Android-first apps succeed. Of course, the fragmentation is a real problem, and there hasn’t been an existence proof of an Android-first app that’s had the same level of traction as, say, Rovio or Instagram.

More platforms upcoming?
Let’s also not count out Windows Mobile, or maybe even a resurgence in native applications as Microsoft and Apple build out their desktop app stores. There’s also interesting emerging companies like Pinterest or Dropbox, which may not be in the 100s of millions of users, but may quickly get there.

I predict that marketing channels will loosen up in the short-term
Lots of interesting choices here – there’s a ton of opportunity and I think we’ll see that the competition between platforms will lead to a loosening of distribution channels. Facebook will hopefully open up a bit more, and provide a bunch more traffic, rather than see all their social gaming developers sucked into mobile, for instance. Will be great to see.

Like this post?
Get new updates via newsletter..

http://twitter.com/sachinag Sachin Agarwal

I’m curious to know if you think that Web and e-mail don’t count as legitimate platforms to build on any more. (Obviously in asking the question, I do, so would love to hear your thoughts.)
http://earnedmedia.wordpress.com/ Christian Brucculeri

I see email as a channel – it’s not really a platform.

The web’s platform is basically Google, which provides ‘distribution’ through SEO and SEM
http://www.facebook.com/andrewaddodito Andrew Chooah

A good read. It’s easy to count out Windows mobile at the moment, but when Windows 8 drops and Microsoft invests heavily in coupling, we might see some market shift take place.
http://petegrif.tumblr.com/ Pete Griffiths

One small quibble and a question:
a) quibble
you don’t mention one of the most important platforms of the last 30 years – UNIX/Linux. I appreciate it isn’t a consumer facing platform but as it clawed its way to its currently dominant position in infrastructure it faced many of the same issues and many of the lessons learned by those who took part in that struggle apply to this day.
b) question
Could you please elaborate on what you mean by marketing channels ‘loosening up?’
http://andrewchenblog.com Andrew Chen

Good point re: UNIX/Linux. Since I’m so focused on consumer, I don’t think of it very much as a viable platform in itself except as it is packaged into “desktop software”

Re: marketing channels loosening up, I mean that Facebook, Apple, and others will try to help developers get more distribution for their apps rather than either reducing distribution or even worse, competing with their app developers.
http://andrewchenblog.com Andrew Chen

Agreed, and this relevant link even came out today:
http://blog.flurry.com/bid/86277/Microsoft-May-Be-Closer-Than-It-Appears-in-Android-s-Rearview-Mirror
http://andrewchenblog.com Andrew Chen

Re: email, the performance of emailing invites as a means for distribution has severely degraded over the last couple years, due to overuse of this mechanism (I wrote about this in Law of Shitty Clickthroughs a while back). To build a mailing list, we saw Groupon and others buy Facebook and other remnant ads, rather than asking people to import their addressbooks, for instance. We’re long past the point where it’s a viable platform to distribute a product.

Re: web, I should have added more detail to my “Facebook” answer to include web. After some number of years, we’re now at the point where once you have a website you really have two realistic options:

1) Integrate into a social platform push content into it to get traffic back
2) Pay Google

Given that, web as a standlone without any social or mobile hooks is no longer a very competitive solution to getting distribution.

It’s hard to speak in generalities about this because I’m sure you could still make a breakout that defies the odds described above, but I think it’s certainly a lot harder and the productive channels have moved to social and mobile now.
http://petegrif.tumblr.com/ Pete Griffiths

My observations of the behavior of platform companies is that the big problem that would be developers face are generally not so much competition from the platform company (it generally isn’t too hard to read which areas they are likely to extend their reach into) so much as the way that the platform companies favor existing large developers. They are so symbiotically enmeshed with these companies and their resources to help other developers are so limited that they end up giving only nominal support to fledgling developers. A good example is Zynga on FB. It is a struggle for small companies to break out because all the traffic is hogged by Zynga and that suits FB just fine. They can throw a bone at small developers from time to time but that is pretty much it. But is that ‘more distribution?’ What would more distribution really mean in these circumstances? Better discovery? That would certainly help because most app stores suck. But it isn’t the platform app stores that are trying to solve the problem – it is…other developers. My feeling is that it is wise for developers to have a very jaundiced view of the help they are likely to get from the platform companies. They are typically useless until you really really really matter to them and then you don’t need them.
efeamadasun

Great article. But, I feel the Chrome webstore should have been mentioned. I feel it has a lot of potential, and between chrome browser, chrome book and chrome box, they have the channels to distribute apps on a really large scale.
http://andrewchenblog.com Andrew Chen

I think they need a bit more of a track record before they can go on this list. G+ as well.

@andrewchen