I Didn’t Know I Was Running a Behavioural Experiment

Part 2 of 3: Rethinking the Exit Interview

In Part 1, I told you what I did. What I didn’t tell you — because I genuinely didn’t know at the time — is why it worked.

I wasn’t running a behavioural experiment. I was just tired of getting useless answers. It was only later, when I started reading Richard Thaler’s Nudge, that I found the language for what I’d stumbled into. Thaler’s central argument is simple: the way a situation is designed shapes what people do in it — often more than any instruction, policy, or incentive ever could. Small, invisible design choices nudge people toward certain behaviours without forcing them. Sound familiar?

What I’d built — without quite realising it — was a nudge. Several of them, actually. And once I understood nudge theory, I understood exactly why my exit interview conversations produced what they did — and why the traditional ones don’t.

Choice Architecture

This is the big one. Thaler defines choice architecture as the design of the environment in which decisions are made. Most exit interview processes have terrible choice architecture — built accidentally to produce exactly the kind of safe, diplomatic answers that tell you nothing useful about why employees leave. I redesigned the environment. Split the conversation. Changed the room. Everything else followed from that.

The Audience Effect

People behave differently when they think something is being recorded. It’s one of the most reliable patterns in behavioural science — and it quietly ruins most offboarding conversations. The moment I said “nothing from here goes on record” and showed them the screen as proof, the psychological context shifted entirely. They stopped performing. They started talking.

Autonomy as a Trust Signal

Thaler is big on what he calls libertarian paternalism — nudging people toward better outcomes while preserving their freedom to choose. I applied that here. Share as much or as little as you want. Giving people control over how much they shared paradoxically made them share more. Forced sharing creates resistance. Chosen sharing creates openness. It’s one of the most underused tools in people management.

The Likert Scale as an Anchor

The six rating questions weren’t measuring anything. They were starting conversations. Thaler writes extensively about anchoring — how an initial reference point shapes everything that follows. A 3 on a 1–5 scale is a natural, non-confrontational hook. Far easier to probe than a blank page. It gave us both something concrete to look at, which made the whole thing feel less like an interrogation and more like a genuine employee experience conversation.

Regret Framing

This one I’m most proud of figuring out intuitively. Complaints implicate others. Regrets are self-referential — they don’t point fingers, they just reflect. What do you wish had been different is a fundamentally different question to what went wrong. Nobody burns a bridge by admitting they wish they’d spoken up sooner. Regret is the side door to the truth — and it almost always opened into something far more useful than any employee engagement survey ever surfaced.

Prosocial Motivation

Thaler’s research on social norms shows that people respond powerfully to the idea of helping others. Help us help your friends bypassed self-interest entirely. People will do for others what they won’t do for themselves. In a conversation about employee turnover, giving someone a reason to be honest that’s bigger than themselves is one of the most underused levers available to HR professionals and leaders alike.

Pre-Existing Trust as the Foundation

Here’s the one no nudge can manufacture. None of the above worked in isolation. The audience effect, the autonomy signal, the regret framing, the anchor, the prosocial appeal — all of it sat on a foundation of trust built long before anyone handed in their notice. Psychological safety isn’t created in an exit interview. It lives in the workplace culture and organisational culture long before someone decides to leave. What I was doing was creating the conditions for truth to surface. But the truth only came because the organisation had already earned the right to hear it.

That’s what most people strategy frameworks miss. The exit conversation reveals what was always already there — the team dynamics, the unspoken tensions, the moments that shaped someone’s decision to leave long before they handed in their notice. It just finally makes it visible. And for any HR leader serious about reducing employee turnover, that visibility is everything.

In Part 3, I’ll make the case that the traditional exit interview doesn’t need improving. It needs replacing entirely.