Recently one of our clients came to us all distraught because of the announced Parse shutdown. Like so many of their 100,000 counterpart developers who relied on the Parse backend for their mobile infrastructure needs, they also used Parse in their offering.  They were left scrambling when Facebook announced its decision to discontinue Parse, a mobile app development platform. So whether or not you have been burnt by the Parse Fiasco, there are quite a few lessons to be learnt here, regardless of whether you are a CIO or an architect.  If you want to make sure to that your business makes sound business investments, continue reading…

The Parse Background

So what is Parse (or was)? Parse offered a lot of benefits for developers, including being extremely cheap and easy to use, with little need for server backend knowledge by the developer. Parse was a mobile app development platform with a system of databases that managed operating systems and tracked user information. Apps could have features such as push notifications easily. It had high scalability, and included analytics and storage.

Facebook acquired Parse in 2013 for a reported $85 million with an estimated 100,000 apps being supported at the time.  To Facebook’s credit, they spent time and money investing in Parse over the last three years, adding features to its service and tools to make it even more useful.

Sounds amazing, right? And herein lies the rub – just as the developers got more comfortable with the platform, Facebook had other designs.  So everyone is in the doldrums, now that they are suddenly expected to migrate elsewhere.

(Slightly) Good News

Parse is helping their users migrate to other platforms, but apps and their features can stop working. Facebook has open sourced the Parse platform with help for developers to install on their servers.  However, the irony of this situation is the whole promise of not needing to know the backend becomes oxymoron when you expect those developers to do the heavy lifting of porting the whole complex environment elsewhere, adding countless hours of work, and plenty of headaches and frustration.

However, the availability of the Open source code and the instructions does provide a silver lining to the dark Parse Cloud situation.

The Solution

So, we decided to work with our client to clone a Parse server and test the theory of migration.  The good news is – it works.  We are helping to migrate in steps by reducing the independence on the third party code in phased manner.  Eventually, we may decided to leave the cloned Parse server as a part of the client’s offering.  We will write the details of the experience in another entry soon.

In the meanwhile, there are 2 major takeaway points in all of this:

1. Select a mobile development outfit that will work with you on a long term basis.  If the company has a long standing experience in developing and supporting real-life applications, then you will get the benefit of the deep experience and sound business and technical decisions.
2. Although the lure of a highly touted backend service may seem attractive, think about the incremental development of your backend strategy.  In many instances, you need only a fraction of the services to start with your MVP and you can add the features and services as you iterate or pivot your offering.  Once again, a company or partner that has an experience in end-to-end technology stack as well as real-world business savvy will keep your development costs lean while keeping the destiny of your business in your own hands!

(Note: while the above 2 points are no-brainers, there are times when the expediency of decision will make them seem less important.  Don’t fall victim to this, and any time you are in that situation – feel free to “Parse” your memory archives for this nightmare!)