Alex Weinstein
On Growth, Product, and Leadership
Platform as a Cop Out

It’s fascinating how the Microsoft glory isn’t letting people sleep well at night. Most entrepreneurs I talk to these days understand that the way to create a lasting advantage is to create a platform – something that entrenches you way better than an application.

Quoting Mr. Ballmer of Microsoft, “Developers! DEVELOPERS!! DEVELOPERS!!!

The reason, of course, is that every application that gets built on your platform creates a sunk cost and deeper entrenchment. When a developer decides to take on your platform – when another company essentially BETS THEIR LIFE SUPPLY on your platform, you have an ally. That ally is selling their product to end customers. Every time they make a sale, you win tremendously – even if you didn’t make a penny from that transaction. The entrenchment of your platform – and thus, your lasting advantage – has just gone up, because it’s now more difficult to switch – all that money spent on the existing application is already gone. Why reinvent the wheel? We already have a VB-based Access 97 application that solves our problem..

You’ll laugh and point fingers at IT departments of companies stuck in 1997, but this logic very much makes sense. I’m Johnson & Johnson. IT is very much a SERVICE, not the core of my business. IT Department built applications that do the job. Fine, they aren’t great – they don’t scale or deploy easily, and there’s recurring cost to maintaining them that keeps going up – but compare that cost to switching to “free” Linux or super-user-friendly-and-sexy Mac OS, and the choice is clear. They’re sticking with Windows. Joel Spolsky describes this phenomenon well in his “API War” article.

Here’s a problem, though: everyone understands how great the platform advantage is, but few know what it takes to get a platform to win. They think “we have a sexy product; let’s now open up our product for various partners to build on top of! Come on, biz dev people! GET US SOME PARTNERS ALREADY!”

Ummm. How do I say this politely. Your platform story is a cop-out. A platform is viable only if it comes with several MAGICAL applications built on top of it. Applications that INSPIRE developers to play around and experiment. Applications that demonstrate the power and desirability of the platform.

Let’s look back at what Microsoft did when pushing Windows. The key to that victorywas MS OFFICE. Office is the flagship demo application that has always shown developers that OH MY GOD this platform can really be used to build amazing shit. When MS was pushing Windows over DOS, nobody really knew what this Graphical User Interface thing was about. Nobody trusted it. And KABOOM, here comes a fantastic suite that everybody wants. The application DRAGS THE PLATFORM by its ears towards success. Suddenly everybody wants to get Office – but to run it, you need Windows. More users have Windows -> more developers feel like creating software for it.

Another example: Apple and its iPhone development platform. Look back at iPhone v1. It had NO custom applications. Why? Because it made no sense to invest into the platform before there was a flagship set of products to demonstrate its power. As soon as end-users LOVED the mail app, developers took that app as a paradigm for stacked panel interfaces. Customers ENJOYED the tabbed music player interface – and developers started imitating it in their apps, too. If Apple took the Microsoft mobile phone road with its software back then – clunky built-in apps, no real flagship end-to-end scenarios to imitate – iPhone would have been a flop just like Windows Mobile 6.

Here’s the rub: nobody knows your platform better than you do. If YOU can’t ship a few amazing end-to-end solutions on top of your platform, how do you expect others to do it? External partners will start with a little tinkering – replace the MySQL adapter in your sample app with a SQL Server adapter. Maybe change the reporting engine from the sample MS Excel to Tableau. Oh wow, it still worked. I guess I’ll be brave and try a bolder modification of the sample app now.

If you’re building a platform – or even harder, a platform company – make sure that you have some SHINY end-to-end solutions built on top of it the day you ship.