We tried to automate making coffee and somehow ended up rewriting our entire onboarding process

In This Story

It started, as these things do, with a complaint. Specifically, somebody said the coffee maker was “basically a paperweight with ambition,” and somebody else said “we should automate it,” and then before anyone could stop us, a Trello board was open and feelings were being typed into a shared document. This is how every Paper Plate Labs experiment begins. It is also how most of them go sideways.

The original idea was simple. We would build a workflow that detected when the coffee pot was empty and pinged whoever was closest to the kitchen. Easy. Twenty minutes, tops. We had a smart plug. We had a webhook. We had ambition. What we did not have, in retrospect, was self-awareness.

By hour two, we were debating who counted as “closest to the kitchen.” By hour three, we had built a small leaderboard tracking who was making the most coffee. By hour four, somebody pointed out that the leaderboard was, technically, a performance review. By hour five, we had accidentally redesigned our onboarding process around coffee equity. None of this was on the roadmap. None of this was approved. All of it shipped.

What we learned

The technical lesson was modest: webhook chains are a delight until they are not, and you should always log the thing you think you will never need to log. The bigger lesson was that small experiments have a way of leaking into bigger questions. A coffee bot turned into a fairness conversation. A fairness conversation turned into an onboarding doc. An onboarding doc turned into a Tuesday.

We did not solve the coffee problem. The pot still gets left empty. Greg still attends. But we did learn that our onboarding was missing a section called “unwritten kitchen rules,” which now lives in the handbook between “how to request PTO” and “do not microwave fish, ever.” Both are equally enforced.

Things we built along the way

  • A smart-plug script that reports coffee pot status to a Slack channel that nobody reads on purpose
  • A leaderboard nobody asked for but everybody now checks
  • An onboarding doc that is somehow now our most-read internal page
  • A small grudge between two team members that we will not get into here

Would we do it again

Yes. Immediately. With less planning, more snacks, and the same poor judgment that got us here. That is the lab, baby. If you want to try this yourself, our only advice is to commit fully and document everything, because the experiment is never the experiment. The experiment is what happens around the experiment. The coffee is just the coffee.

If you have built something similarly dumb and accidentally useful, we want to hear about it. We collect these like other people collect coins, except ours are mostly free and slightly stained.

Written by Daniel Burns

For over 20 years, Daniel has a proven record of managing strategic direction and development – driving revenue gains and improving their efficiency and effectiveness. Daniel grew up in Brazil and graduated from the University of Miami. An avid reader and adept at radical sports, he sailed across the Atlantic and biked across Europe. These broadening experiences give him the ability to think laterally, turning stakeholder expectations into successful digital strategies.