"The team almost always knows a project is dead before the leader does." — @JoshWoodward (Head of Google Labs, Gemini App, Google AI Studio)

@joshwoodward Then the leader's job is to make it safe to say so early. Ask the team directly in every review whether they still believe in it.

@joshwoodward The team knows first because they see the usage logs and workarounds every day. Leaders see the roadmap. The fix is putting raw user sessions in the leader's week, not summaries.

@joshwoodward this is so true it hurts by the time leadership admits a project is dead the team has already started mourning it quietly

@joshwoodward Give the team a landing spot when a project ends. If admitting it's dead also means 'my role disappears,' you're asking people to vote against themselves. Moving good builders onto the next problem makes stopping a project less of a personal loss.

@joshwoodward so how does the team actually tell the leader tho

@joshwoodward Painfully true. I’ve been on both sides: the quiet “this is dead” hallway talk, and the leader still polishing slides. Persistence isn’t the virtue when reality already voted. The skill is hearing the room before your ego does.

@joshwoodward the fix that works is making it cheap to say out loud, one line in the weekly update on what would make us kill this. if saying it is career risk the team stays quiet

@joshwoodward leaders are usually the last to hear it cuz nobody wants to be the one to say it

@joshwoodward Usually the team does know, and they keep quiet because saying it out loud has cost someone before. A leader who hears it late should look at what happened to the last person who said it early.

@joshwoodward painful truth

@joshwoodward This is one of the cleaner uses of AI for a leader: feed it the raw signals from the team and the dead projects surface while there is still time to act. The analysis was never the bottleneck, wanting to hear it was.

@joshwoodward The leader usually isn't the last to know. They're the last to admit it, because their name is on it. So everyone waits for a face-saving exit: a budget cut, a reorg, "new priorities." The project dies six months late and nobody gets blamed.

@joshwoodward always. nobody wants to be the one saying it out loud

@joshwoodward Ask the team every month, anonymously, if they'd still start this project today.

@joshwoodward For AI pilots there's a cheap way for the leader to find out sooner: read the usage logs. When the team that built the pilot stops opening it, the status deck stays green for a while longer. I'd ask for weekly active use by the builders before any steering update.

@joshwoodward Same thing shows up in GTM: reps know a market is dead when deals only close with a discount, months before the dashboard turns red. For overseas launches I'd write the kill criteria on day one, with the team, so the leader isn't the last to read the signal.

@joshwoodward Dead projects live on in meetings. Demos kill them fast.

@joshwoodward Most AI projects die in the gap between what the builders already know and what leadership is still celebrating. The teams that ship treat a quiet kill decision as a weekly habit, not a crisis.

@joshwoodward Absolutely fantastic interview. Truly enjoyed it. Thank you both.

@joshwoodward same with deals, the rep knows it's dead weeks before the forecast does and the CRM still shows it open

@joshwoodward If the team already knows it's dead, the meeting is just theater for the org chart.

@joshwoodward It's how the uncomfortable conversations are carried out with the leader says about how good a leader actually is.

I'd have the leader name what would make them stop funding it.

@joshwoodward Looking forward to the day that Google is back to making videos about its successes rather than its failures.