Shopify’s AI peer-review question was never really about the score. That’s the takeaway from CEO Tobi Lütke’s latest comments.
Speaking with Alex Heath on the Sources podcast in an episode published Oct. 28 and recorded at Shopify’s Toronto office, Lütke reflected on the company’s April 2025 memo that told employees “reflexive AI usage is now a baseline expectation at Shopify.”
He said the point of adding AI usage questions to performance and peer reviews was that having the question there can “give a company a reputation to live up to.” The actual peer-rating answer, he added, didn’t matter.
The metric that mattered by existing
Lütke said he no longer has to enforce rules: usage and headcount justification now happen on their own. The memo drew more attention outside Shopify than inside, he noted, and he read that as a sign the transition had already happened.
He said he doesn’t know whether the AI question is still in Shopify’s current review. The question asked peers how reflexively a colleague shows up with an AI prototype, and he admitted “sometimes you play a little bit of tricks there.” But the answer itself wasn’t the point.
For marketing and growth teams, this is a useful reframe. A metric can shape behavior simply by existing, because it makes a norm visible.
- It puts AI adoption into peer conversations, not just manager goals.
- It creates social proof: teammates see experimentation as expected.
- It changes resource debates, since teams must explain why AI can’t do the work first.
AI skeptics and the top-performer question
Lütke went further on skeptics. In a company where everyone around you works with AI agents, he said, a skeptic who tries to lay low will never become a top performer. He mentioned open-source projects that have banned AI contributions and debates over whether code was “hand chiseled by humans or somehow computed by math.” His view: good code should be accepted either way.
He also argued that companies should decide who they’re for and who they aren’t, and that “heightening the misalignment with the people you don’t want is actually almost more important than having a good careers page.” Besides, he said, everyone deserves a place where they can be a top performer, and Shopify wants to help people find that elsewhere when it isn’t possible.
Don’t confuse adoption with quality
There’s a warning buried in the conversation. In a Knowledge Project episode released Sept. 15, Lütke said AI had made one kind of lazy work worse. Shopify staff call unread AI output passed to a colleague a “slop grenade.”
A review question about adoption won’t tell a team lead whether anyone actually checks an AI draft before it reaches an editor or a client. That’s the operational gap for agencies and D2C teams: adoption and quality are two different signals.
Marketers can mirror Shopify’s play without copying its headaches. Add an AI-behavior question to peer reviews to signal the norm, but pair it with a separate quality gate: read it, verify it, then send it.
What happens next
Lütke said he expects public pessimism about AI to fade as more people use it themselves, and that stories from that use will wash it away. He didn’t say whether the question will appear in Shopify’s next review cycle.
For now, the lesson is less about a scorecard than about culture. The question was the intervention.
Source: Search Engine Journal
