Would you build a business around public APIs?
- Would you use currently available APIs (including those on marketplaces) for building a business?
- What would be your criteria of a suitable API to use it in building a business' technology?
- What main problems do you see with the current API ecosystem?
Sometimes the cost (money, time, effort) of building your own API/solution exceeds the cost of using an existing solution. If this is the case, you should consider using the existing, commercially available solution.
Thanks @ElijahGeorge! What would be your advice for picking the right API out of several seemingly equivalent alternatives? What to look at?
Absolutely! Lots of people do that to build out a SaaS at lower initial cost, or to vet a SaaS solution. You may even be able to scale to a decent size. At some point, when you realize that have some traction, you can build out the pieces and replace the public APIs over time. Is it a perfect solution? Certainly not, but "perfect" costs a lot of money and it's best to confirm you have a brilliant idea for a product before you built it.
The main problem I see is that integrations can sometimes be temperamental. If I were doing it, I'd build code around the integration points to add fault tolerance and resilience. I can help with that if you need help. I'd first wire up the pieces you need though and see where the problems show up.
Thanks for your thoughts @iamericlentz! Do I get it right that you only see APIs as temporary solution for speeding up the launch? In that case, what is that mysterious growing API economy a lot of people write and talk about in recent years? Especially taking into accounts that a huge number of APIs available now are simple directories and collections of facts like, a list of countries, phone codes, exchange rates, etc.
Are there any examples of serious APIs carrying out true business tasks, being used on a permanent basis?
I wonder if I'm not 100% on the same page? Maybe you meant something like this example?:
I built a SaaS for a client that presents maps to trail riders. We use an API service that provides the map data. That's all they provide. They don't offer applications that use that data (as far as I know).
That's sufficiently complex that I would never suggest recreating it. In that case, we're happy to use that forever, especially since we haven't had issues with it.
My general thought is that if the API isn't offering a considerably complex solution (e.g., list of countries), then just build it yourself, either sooner or later, to minimize dependencies and fault points and perhaps enhance performance (caching notwithstanding).
An API economy can be supported by those who just aren't motivated to build something similar. The use case I'm promoting - short term while getting to needing to scale - could likely support such an economy. Then, there are those who will never replace it if it is working well.
No. At least not in cases where the API is responsible for the main functionality of my app. I had one of my apps killed years ago because a business decided it was more profitable to close their API to other businesses and thus locked everyone out over night. Months of work wasted.
Thanks for sharing your experience @nikwen! Sad to read that. Was the API unique? Could it have been replaced with some alternative?
There was no alternative and developing one would have taken years, even with a team. F**k you, Tapatalk!