Developer Tools
Parse
Parse was a backend-as-a-service that let mobile developers skip building their own servers; Facebook bought it in 2013 and roughly 600,000 apps came to depend on it. Then Facebook decided the developer-platform business wasn't core, and shut Parse down — a landmark lesson in the risk of building on someone else's platform.
- Company
- Parse
- Started
- 2011
- Ended
- 2017
- Apps built on the platform
- ~600,000
- Collapse speed
- Gradual
- Preventability
- Low
- Lesson transfer
- Universal
- Last reviewed
- 2026-07-22
Narrative
The story
The ambition
Parse promised developers freedom from the back end. A backend-as-a-service, it gave mobile developers ready-made SDKs and APIs for the unglamorous plumbing — data storage, user accounts, push notifications — so they could ship an app without building and running their own servers. For small teams, students, and solo developers especially, it was foundational infrastructure. Facebook liked it enough to buy it in 2013 for a reported ~$85 million and make it a flagship developer service.
The rise
It became genuinely essential. At its peak roughly 600,000 apps were built on Parse — an entire ecosystem of products that ran on infrastructure Parse (and then Facebook) operated on their behalf.
The cracks
But Parse was never core to Facebook, and the market around it changed. Amazon, Google, and Microsoft — along with numerous startups — came to offer very similar developer tools, commoditizing the category. Facebook decided the business was no longer worth maintaining.
The collapse
On January 28, 2016, Facebook announced it would shut Parse down — giving developers a full year to migrate, open-sourcing "Parse Server" so they could self-host much of the API, and providing migration guides. On January 30, 2017, the service went dark for good, disabling the API app by app. Roughly 600,000 apps had to move to a self-hosted backend or another provider — or stop working.
The aftermath
Parse became the reference example of platform dependency: essential infrastructure that vanished when its corporate owner's priorities shifted. The open-sourced Parse Server let some of the ecosystem live on, but self-hosting was precisely the burden Parse had been created to remove.
The lessons
Building on someone else's platform means accepting that they can walk away. Parse was excellent and widely relied upon, yet its fate was decided entirely by an owner for whom it was never central — and when the category commoditized, that owner rationally exited, stranding the ecosystem that had trusted it. Convenience obtained by depending on a platform is borrowed, not owned.
Causal timeline
Failure Anatomy
- 2013
A backend, as a service
Parse gave mobile developers ready-made SDKs/APIs so they could skip building their own servers; Facebook acquired it in 2013 for a reported ~$85M and made it a flagship developer service. [1]
- 2015
600,000 apps depend on it
At its peak roughly 600,000 apps were built on Parse — an ecosystem running on infrastructure Facebook operated on their behalf. [2]
Platform dependency - 2016-01
Facebook decides to exit
With backend-as-a-service commoditized by AWS, Google, and Microsoft and never core to Facebook, Facebook announced Parse's shutdown in January 2016. [3]
Stronger competitor - 2017-01
Lights out
After a one-year wind-down (with open-sourced Parse Server for migration), Parse shut down on January 30, 2017, forcing ~600,000 apps to migrate or stop working. [4]
Platform dependency
Structured analysis
What Went Wrong
Root causes
A commoditized, non-core business. Backend-as-a-service was commoditized by Amazon, Google, Microsoft, and others, and Facebook — for whom Parse was never core — chose to exit the developer-platform business. [3]
Contributing factors
An ecosystem on borrowed infrastructure. Roughly 600,000 apps had built on infrastructure owned by a company with no obligation to keep it running. [2]
Immediate trigger
Facebook announces the shutdown. In January 2016 Facebook announced it would shut Parse down, giving developers a one-year migration window. [3]
Visible symptoms
600,000 apps forced to migrate. The shutdown forced roughly 600,000 apps to move to a self-hosted backend or another provider, or stop working. [2] [4]
Warning signs
Built on infrastructure they didn't own. Roughly 600,000 apps depended on infrastructure owned by Facebook, which had no obligation to keep it running. [2]
Affected groups
Evidence
Claims & sources
Every numbered marker in the analysis links to the claim it rests on, and each claim to its sources.
- [1]
Parse was a backend-as-a-service that gave mobile developers ready-made SDKs and APIs so they could ship apps without building or running their own servers; Facebook acquired it in 2013 for a reported ~$85 million and made it a flagship developer service.
- [2]
At its peak roughly 600,000 apps were built on Parse.
- [3]
Backend-as-a-service was commoditized by Amazon, Google, Microsoft, and others, and Facebook — for whom Parse was never core — decided the business was no longer worth maintaining and announced its shutdown in January 2016.
- [4]
Facebook gave developers a one-year migration window and open-sourced Parse Server, then shut the platform down on January 30, 2017, forcing roughly 600,000 apps to migrate or stop working.
Sources
Facebook Shutters Its Parse Developer Platform
TechCrunch · 2016-01-28
Facebook's Parse developer platform is shutting down today
TechCrunch · 2017-01-30