The day-31 handoff

On the last day of every engagement we run, we spend the whole day training the client's team to run the thing we just built. Not a Loom, not a doc dump, not a "reach out if you have questions." A real day, with the people who will own it after we leave, using their real accounts, breaking things and fixing them while we're still in the room. Almost no one else does this. Five agencies that overlap with our service mix. Every one of them sells "post-launch support" as a monthly retainer — buy more hours, keep us on the hook. Two of them sell training as a product they'll build for you. None of them will train your team to run the thing they built. That's not a coincidence. Enablement makes you unnecessary. Retainers make you rich. We think the retainer model is wrong for a specific kind of buyer: the small or mid-sized company that hired an outside team because they didn't have one, and can't afford to build a dependency on a vendor to run the thing that runs their business. For that buyer, "we'll always be here" isn't reassuring. It's the problem restated. So on day 31 — the day after the six-week engagement wraps — the client's team owns it. Every credential is in their password manager, not ours. Every deploy runs from their account, not ours. The runbook lives in their wiki. The database backups land in their storage. If we vanished the next morning, nothing would break. Getting there takes deliberate work in the last week of the engagement. The mechanics of the handoff are boring on purpose. A one-page runbook that says exactly what breaks and exactly how to fix it. A recorded walkthrough of every non-obvious decision, tagged by the line of code it explains. A live incident drill — we deliberately break something in a staging environment and the client's team fixes it while we watch. A written service-level agreement with themselves — what they'll respond to in an hour, what can wait until Monday, what they'll pay someone else to handle. The last piece matters most, and it's the one most agencies skip. A handoff isn't just "here's the code." It's a decision the client makes about which parts of running this system are theirs, which parts they'll outsource permanently, and which parts they never should have owned in the first place. We help them make that decision on day 31, not discover it in month four when something breaks at 11pm. There is a version of consulting where the goal is to make yourself the answer to every future question. There is another version where the goal is to make the client the answer. The first one pays better in the short run. The second one is why our clients refer us. If your vendor's proposal has a big "ongoing support" line and no handoff plan, ask them what happens on day 31. If the answer is "you'd stay on our support plan," they're not selling you a build. They're selling you a subscription to their team.