A few months back I wrote about Facebook’s acquisition of Parse. Parse provided a “Mobile Back-End as a Service” (MBaS), which is an ungainly acronym for what is essentially a cloud-based host for mobile apps. The idea behind Parse and its competitor StackMob, was that mucking about with servers can be cumbersome, and by using their services developers could focus on the more important parts of their own job – designing app mechanics and user interfaces (UI). What these companies offered was a bit more complicated than Amazon’s AWS cloud computing service. Parse provided a set of tools for building, deploying and managing apps – sort of a suite of applications that run on top of the ‘operating system’ provided by Amazon. This is known as “Platform as a Service” (PaaS) as opposed to Amazon’s “Infrastructure as a Service” (IaaS).
Then in December PayPal announced it was acquiring StackMob, and this week announced that they were shutting down StackMob’s service to focus the team on other projects. Facebook appears to have bought Parse to boost tools for mobile developers to use in conjunction with Facebook’s core sites, but still operates the basic Parse service. By contrast, PayPal appears to have just hired a team of mobile-saavy developers.
The idea behind MBaS, on its face, seemed to make a lot of sense. So what went wrong?
First, it seems likely that AWS and other cloud providers already offer a simple enough set of tools that app developers did not really need the extra services the MBaS providers offered. Maybe mucking about with servers is not as cumbersome as we thought. Second, once you start talking about developer tools, you quickly start to cover a near infinite set of features. Niche service providers started appearing that handled specific developer tools, and no single MBaS provider could match all of those specialists. For instance, Urban Airship, which handles notifications and geo-location, and Crittercism which handles error monitoring, both out-competed the MBaS’ comparable services. Developers are a highly varied group with a wide range of skill sets, so they apparently prefer being able to order off the menu a la carte. The MBaS generalists were thus stuck in between AWS nearly infinite set of basic tools and the narrow specialists.
As far as I can tell, both Parse and StackMob got fair-sized exits, so this story has a happy ending. But I also think there is a broader story here. As stuffy as the app market is starting to feel there is still a lot of innovation taking place out there. The industry has room for new services, and the model will continue to evolve. In my original post on Parse last year, I pointed out that “Developer Tools” was long seen as the graveyard of business models. The 1990’s saw dozens if not hundreds of these companies come on stream, but all of them eventually shut down or got acquired. In the world of shrink-wrapped software sold on physical discs, there were not enough developers to sustain a specialized tools industry. The advent of cloud computing and wildly open distribution platforms (aka App Stores) has dramatically lowered the barriers to entry to software development, and rekindled the market for developer tools. Unlike their ancestors from ten years ago, today’s tool providers charge tiny amounts (Freemium in many cases) and hope to make that up in volume. I think there is still considerable room for change here, and merits continued attention.
Another possible reason: mobile apps are getting complicated enough for an app developer to need tech staff, whether outsourced, or in-house. Create-Publish-Forget-about-it doesn’t work any more with apps, like it did in the early days. Once you have a tech team, datastore + server side logic is the easy part – a simple LAMP stack on AWS is enough. The gap then is in the specialized components that are not already built out for the web. Where these SaaS companies like Urban Airship are able to step in.
I think we’re yet to see the WordPress / Squarespace / etc. equivalents for mobile.
Makes sense. LAMP on AWS now seems almost ‘old’, and very possibly more expensive than doing it yourself.