June 3, 2026
Hi Everyone,
The hardest thing to spot in a product team is the moment it stops learning. The team keeps shipping new features, but the question of whether each feature is worth it stops getting asked - and a few quarters go by before it shows up anywhere a leader can see it.
Today we're walking you through six early signs of a product team that's stopped learning, with a reset move for each.
1. Nothing has been removed in the last 12 months
A team that never removes anything ends up paying to maintain work that customers don’t value (or use!), while the engineers who should be building the next thing spend their weeks keeping the old thing alive.
Ask your head of product to name everything the team killed or sunset in the last 12 months. The length and confidence of the answer tells you where the team sits.
Pick one feature that's expensive to support and used by fewer than 5% of customers. Plan a clean sunset, with a real communication and migration plan for the customers who do use it.
2. The team can't tell you what worked in the last launch
A team that can't tell you whether its last launch worked has nothing to learn from. The work keeps coming, but the lessons that should have shaped the next round never arrive.
Ask your head of product what the last meaningful thing the team shipped was, and how it performed. A strong answer covers adoption, a customer story, and one thing the team learned.
Shopify uses a product process with five phases, the last of which is called Results. Every project ends with a structured review of what the launch actually delivered. Add a Results step to your own cycle, with the same person who announced the launch writing a short note on the adoption number, one customer reaction, and one lesson worth carrying forward.
3. Outputs within your OKRs
Look at your team's current objectives. If a key result reads like "Launch onboarding v2 by end of Q2," the team will deliver the output, regardless of whether v2 made any difference to new users – they committed to building the thing, not to changing the outcome the thing is supposed to drive.
A real key result names a number the business cares about and a level it has to reach. "Bring new-user activation from 38% to 55% by the end of Q2" puts the team on the hook for the outcome and leaves them free to argue about which feature, or no feature, gets them there. Perhaps launching a v2 of onboarding suddenly doesn’t even make the list.
This week, pick one key result on the current quarter and rewrite it so it states a desired outcome, not an output that may or may not influence the outcome.
4. The same fires keep coming back
Every team gets hit with fires – production goes down, a big customer escalates, sales asks for something urgent on a Friday afternoon. None of that is a sign the team is in trouble.
The warning sign is fighting the same fire for the second or third time. Nobody fixed the cause after the first one because there was no slot in the cadence for that kind of work, so the problem keeps coming back, and the team's calendar fills with reruns of problems it already solved once.
At your next leadership review, list the incidents that ate the most time in the last 90 days. For each one, ask what would have to change for that fire to never come back. Fund those changes before you fund anything new.
5. People are being moved between projects every cycle
John Cutler, who spent years studying product teams at Amplitude and Pendo, calls this "team tetris." Engineers and designers get shuffled from one project to the next, never long enough to learn enough about the customer they’re serving or the system they're working in.
A group that owns the same problem for two cycles in a row will ship something, see what happened, and improve on it. A group that gets reshuffled every quarter will only ever do the first part.
Look at your last two planning rounds. Count how many people stayed on the same problem across both. If fewer than half did, the team is being moved around too often to learn from its own work.
6. The wins are celebrated before the results are known
Watch what gets a Slack post. If launch day gets a huge announcement and a big round of applause, but the 30-day adoption check passes without a peep, the team is being rewarded for shipping rather than for the thing shipping was supposed to do.
What gets celebrated is what the team learns to chase. A team rewarded for outcomes will get sharper at picking what matters most.
For the next launch, hold back the public celebration until the 30-day numbers come in. Then post both the launch and the result together.
Go deeper
👉 John Cutler: 12 Signs You're Working in a Feature Factory – the original list of warning signs that shaped how a generation of product leaders thinks about team health.
👉 Pendo: The 2019 Feature Adoption Report – the source of the 80% figure, with the full methodology and the implications for where engineering money is going.
👉 Lenny Rachitsky: How Shopify builds product – Glen Coates, Shopify's VP of Product, on the five-phase GSD process and why the Results review is the part most teams skip.
Coming up tomorrow
Tomorrow, we're walking you through a single question that tells you whether your executive team functions as one.
That's it for today!
P.S. Out of the six signs, which one hit closest to home?