The First Pancake
Everyone who joins my team is required to push something to production within three days. Some people freeze up upon hearing this on their first day, either from imagining the amount of work they’ll need to squeeze into the next 72 hours or from thinking I’ve lost my mind. Before they convince themselves of either case, I explain my intentions:
“The first pancake always sucks,” I say. “So let’s just get it out of the way.” They ease up and get going.
This has several benefits:
- It immediately sets the pace expected of them, to ship stuff quickly and imperfectly.
- It gives them ownership and responsibility from day one, which is often why they joined a startup in the first place.
- Allows them to charge the trust battery early — to earn trust on top of what they started with.
- Sidesteps the internal hesitation one feels when facing a blank canvas or new notebook. Not only is there a deadline to make the first stroke, they also have permission to make it a sloppy one.
- Gives them a quick win to bolster their confidence.
- Gives me a reason to give kudos in front of the whole company.
- Gets the first attempt out of the way quickly so they can move on to the next iteration, which is often heaps betters.
Ultimately, this cuts the time it takes for new people to become productive, trusted, and satisfied members of the team.
And, despite talk of lousy first attempts, the first pancake can be quite tasty after all. For example, Pinecone’s articles about vector embeddings and vector database were first and second pancakes that took top spots for “vector embeddings” and “vector database” searches on Google, and to this day continue to bring in thousands of organic visitors to Pinecone. If we waited to do those projects after some arbitrary onboarding period, they’d take many more weeks for the same results, at best.
“There’s no harm in just accepting that the first one will be bad, eating it there and then, and moving on to the next, better one.” — Henrietta Inman, Author and Pastry Chef
