Psychological safety isn't enough

Hi Everyone,

"Build psychological safety" – that's the standard advice for getting your team to speak up.

But Amy Edmondson, the Harvard professor who developed the idea, has spent a decade pushing back on how it's used. In a 2025 HBR piece, she wrote that when a team says "we're safe because we never argue," that's a red flag.

Today we're breaking down why safety on its own falls short, and one fix Reed Hastings built at Netflix after a big decision went wrong.

Why "safe" teams still go quiet

Edmondson argues that safety is only one half of what a high-performing team needs. The other half is accountability, so having clear standards, specific expectations, and consequences when the team falls short.

Teams with high safety and low standards land in what she calls the "comfort zone." People feel fine, but nobody pushes back, and the hard questions stay quiet.

What you want is high safety paired with high standards. People feel safe enough to speak up, and the standards make sure what they say matters.

Most engagement surveys only cover one side of this by asking whether people feel safe, but rarely whether the team challenges each other or makes hard calls together.

That's how a team can score great on safety and still go silent on the hardest decisions.

What Netflix did after Qwikster

In 2011, Netflix tried to split its DVD-by-mail business into a separate company called Qwikster. Customers hated it. The company reversed within weeks.

Looking back, Reed Hastings concluded the failure wasn't a strategy problem. He told Tim Ferriss in 2024 that he'd been "messianic" about the move. Several executives had serious doubts, but none of them spoke up.

So he started writing a memo for every big decision and sending it to dozens of senior executives. Each one rates the idea on a scale from −10 to +10 and writes one sentence explaining why.

"If we had done that at the time," Hastings said, "we would've seen tons of −7, −6, −8, and that would've been shocking."

It works because the output is a document, not a meeting. Nobody has to challenge their boss in person, and nobody has to be the first to raise a hand. The ratings do that work for everyone at the same time.

How to test this with one decision

You don't need to do this for every decision. Try it once, on something you're debating this month. We've put together a template you can copy — link at the bottom.

1. Pick the decision: Choose one that's still open and where you can name at least two paths. If you've already made up your mind, this won't work.
2. Write the proposal on one page: Our free template has the sections – the question, your recommendation, what could go wrong, and what you don't yet know. Keep it to half a page if you can.
3. Pick 8 to 12 reviewers: Mix seniorities and functions. Include at least one person outside the immediate team. Outsiders see what the team has stopped noticing.
4. Share with a 48-hour deadline: Ask each person to rate the idea from −10 to +10 in the shared document, and write one sentence explaining their number. Without the sentence, a number tells you nothing.
5. Read everything before you decide: Look for the surprises. A −6 from someone you expected at +5 tells you more than five +7s from people who always agree with you.
6. Make the call and share the synthesis: Send everyone who rated a short note covering what you decided and what the most common concern was. People rate honestly next time if they see their input was read.

After a few rounds, you'll start to see who calibrates well and who consistently misses risk. That's information you can't get from an engagement survey.

Go deeper

👉 Harvard Business Review: What People Get Wrong About Psychological Safety – Edmondson and Kerrissey on the six common misconceptions, including the team-that-never-argues red flag.

👉 CNBC: Reed Hastings on the 3-word tactic that helped Netflix grow – how the −10 to +10 rating works in practice, in Hastings' own words.

👉 Amy Edmondson at TEDxHGSE: Building a psychologically safe workplace – a 12-minute introduction to the framework and how it combines with high standards.

Coming up on Monday

On Monday, we're breaking down how to spot, pick, and resource your first AI champion before you spend another cent on tools.

Thanks for spending this week with us! Have a great weekend.

P.S. Know someone who keeps getting blindsided by issues their team "should have raised earlier"? Forward this to them.