What is the best approach to have a Gamification platform ? Build or Buy?  Why?

My immediate thought is that I struggle to see how a successful gamification can be implemented through a 3rd party gamification platform. Also that the core principles of make/buy decisions haven't changed. Disclaimer: I've been a COTS/SaaS implementer for a very long time because imho, most of the time buy beats make on cost optimisation 80% of the time for most things.

But gamification is not like production planning. One produces in order to produce; but one does not gamify in order to gamify. You gamify in order to elicit desired behaviours within a target audience along a target process. And so, a third party gamification platform (or a home-grown separate platform) will definitely provide all of the key mechanics (tracking, badges, rewards, leaderboards, and even APIs) but is the mechanics enough? 

It seems to me observing gamification case studies that where it has succeeded it is because it is core to the actual activity. For example, we have used hackathons gamified for complex problem solving. And the platform for the hackathon is the platform for gamification. We have also gamified some HR processes, but we manually worked the gamification mechanics through the core process and tools. A separate engine might have been useful, but wasn't truly necessary.

If looking strictly at architecturally linked platform, vs core gamification mechanics embedded in the process and its automation solutions, then I would struggle to see how:
- robust, quality, full featured gamification mechanics *built* can come close to the cost and performance of the leading dedicated gamification platforms.

On the other hand, I can very well see that a homegrown basic gamification engine for a specific application is going to be much much cheaper and easier and fast to build and maintain than a full featured gamification platform, as long as we are happy with our home-grown basic gamification champagne. I.e., "dogfooding" can produce some great wine. But the best Champagne comes from France and costs more. Period. Here it entirely rests on the specs. Do you need a central platform with API extensions into all of your ERP or apps, and are you a majorly COTS/SaaS already. Or do you have any army of devs, just itching for the next thing and maintenance and quality of your home-grown platform is under control? Then build the thing.

My conclusion:
For most cases, COTS/SaaS beats Home-grown 80% of the time.
But a gamification platform as an architectural component is, imho, an exotic choice compared to rethinking the process and building game mechanics into it organically.
The value of full featured gamification platform with all of its features is questionable. Because you don't have to gamify the whole of your stack, do you?

What's the true problem you're trying to solve? 

If you're looking to gamefy staff outcomes, then build it is the most cost effective way to go. Only look to build it yourself if it's something you will commercialise outside your organisation - you'll never be able to do it more cost effectively than someone else already has.