Automation projects fail quietly. The build works in the demo, the invoice gets paid, and six months later nobody can change anything without going back to the original developer, who has moved on. The way to avoid that isn't a better contract. It's choosing the right partner in the first place.
Ask who owns the result
The single most important question: when this is done, do you own the source code and the accounts it runs on? If the answer is 'it runs on our platform' or 'we host it for you,' you don't own it, you're renting it, and the rent only goes up. Insist on full source on your own Git and credentials in your own accounts.
What good looks like
- Full source code, documented, handed over, not a black box you can't open.
- Runs on your accounts and keys, so leaving is always an option.
- Plain-language runbooks, so your team or any developer can maintain it.
- Short cycles with demos, so you steer early instead of waiting for a reveal.
Red flags worth walking away from
Proprietary platforms you can't export from. Vague answers about who holds the API keys. No documentation in the quote. A price so low it only makes sense if they plan to keep you dependent. The honest version costs a little more up front and far less over the life of the system.
Put this to work
3 linksRelated reading
AP automation in 2026: what the research says, and what a 20-person business should do about it
New surveys put the share of finance teams still keying invoices by hand at about 2 in 3, and SMB adoption of AI at its highest yet. A read of the 2026 numbers, where they mislead, and a 3-step plan that fits a business without a finance department.
Intake → Model reads → Checks → Review tab → Your Sheet
When automation is worth it (and when it isn't)
Not every manual task should be automated. A simple framework for deciding what to automate first, and what to leave alone.
How often? → Rules-based? → Error cost? → Automate it
Hiring an automation team vs. outsourcing it
The build-vs-buy question for automation isn't about cost. It's about which failure mode your organization can actually tolerate.
Count the weeks → Steady work → Bursty work → Hybrid → Either way
Continues · Got a workflow like this?