Do I build or buy an API?

Is this a core part of your product offering, or is it a stepping stone to get to where you want to go? Often it's cheaper to purchase a solution rather than building and maintaining it yourself. A big difference is that you see the actual bill for a service, whereas employees cost a fixed amount whether they work on verification software or something else.

At a later point you may find that the service does not support everything you want, at that point you know better what you actually need and then you can re-focus your team to develop a replacement service.

KYC involves several steps to:

• Establish customer identity.
• Understand the nature of customers’ activities and qualify that the source of funds is legitimate.
• Assess money laundering risks associated with customers.

That's not a strictly technical question but legal as well. If you have no legal support in-house there's no chance I'll build that.

On the other hand, having that legal support in-house and a good acceptance criteria beforehand I'll check the planning with the team, estimate an approximate deadline and consider the pros and cons with the CFO and the CEO from an honest PoV.

Depending on the seniority of the team it can cost 1 month to be impossible, if the company can and want afford this to avoid a monthly payment by usage (if the company has a client base they can check how much it will cost) and the team has a good seniority it's feasible to do it, why not?

By the way you can pay external APIs just for a single or a pair of those steps (there are some vendors out there) so you may pay just for the critical path while adding robustness to the rest of the project and swap just this building block on a future or never.

Why build when you can buy and use high quality for cheap? APIs will make your life and the life of your user much better. But make sure that you’re picking the right provider.

I went with Passbase ( passbase.com/ ). I’ve been building a micro trading platform for crypto and I needed something slick, simple, and cost-effective way to verify my users (In India, we need every user to have a KYC the moment they are onboarded)

Passbase made this simple. My users got their KYC much faster (and in just 4 clicks). There are for sure many alternatives but then Passbase gets you started in just under 5 minutes.

I would say try them out first for a small cohort of users. If you still feel like you could provide a much better experience, no harm in building one on your own as well.

Definitely buy when handling personal data and the potential for requiring documents and other follow up actions. Ongoing monitoring may also be a requirement. Long story short though the world figured out this capability and it's now becoming commodity. There are hundreds of services you could use like trulioo, onfido, TransUnion etc.

In the long run, you probably don't want to be maintaining and supporting parts of the system which are not part of the core expertise — even if you could probably build it initially without much problem. Of course this varies depending on the size of the challenge. You don't want to "buy" basic building blocks, but you probably do want to "buy" anything that exists in the marketplace that has all the edge cases accounted for already.

There is no perfect answer here, but my gut is telling me that KYC software is something you should probably buy if it's even a question. 😅

Buying is generally cheaper and faster to obtain. Building is expensive, takes time and requires long term maintenance. I guess there are three cases worth it though: you need something slightly different, you have high "volume", it's a niche market milking cows. Also, you can start with the first while building a replacement if you think it's worth it.