How would you validate what to build and why?
First of all we need to define business/user values. I would research following:
- How many of these clients do we have?
- How much share of revenue do we earn from them?
- Is there alternative ways to catch these clients, f.e. manually integration to their ERP?
- Do these clients have an open API of their ERPs?
- What is happening on the market? Do our (direct/indirect) competitors provide this functionality? What are the standards in the market?
- How do these clients solve the problem in other cases at the moment?
Getting answers on these questions could build a common picture about predisposition in our case to create an API on our side. For example, if our clients have an API (for their ERP) with a doc and we can integrate them manually it could be kind of MVP in this situation. Manual integration will allow us to validate some hypotheses on work with largest clients and datas we need to work with to provide the best experience.
How do you plan to roll it out and manage it over time and phases?
Primarily, we need to define the most important data we need to implement in the API. We can obtain it via interview with our clients and research ready-made solutions in the market.
After we get knowledge of what we need to do, there will be a few main milestones:
- Technical research and discovering.
- Architecture design and coordination potential changes with all related departments.
- Developing an API
- Preparing test accounts and playground
- Preparing a doc to the API (we can proceed it in parallel with the previous point)
- Onboarding page/live demo for our clients
In doing so, the more down level decomposition we prepare on the developing step the better and more manageable the process will be.
How do you see the API shaping the Pleo platform and vice versa?