Grace Judson said something on my podcast I keep coming back to. Trust first, especially after mistakes. And rebuilding trust is never one-and-done.
Most leaders I've worked with treat a broken promise like a bug ticket. Something went wrong. You raise it, you say sorry, you close it. Done.
Your team doesn't close it. They keep it open, and they watch you.
The apology is the easy part
Grace is an executive coach who joined me on Corey-osity Unleashed. Her point is simple. You rebuild trust through a habit you keep, not a speech you give.
I agree with her. I also think most of us get the order wrong. We pour our effort into the apology, then we coast. The research says the effort belongs after the apology.
What happens when you say sorry and nothing else
In 2006, three Wharton researchers, Maurice Schweitzer, John Hershey and Eric Bradlow, ran a trust experiment with 262 people. Each participant played seven rounds of a money game with the same partner. Every round, they started with $6. They kept it, or they passed it to the partner, where it tripled to $18. The partner then decided how much to send back.
The researchers rigged the partner. For the first two rounds, it kept everything and returned nothing. From round three on, it returned $9 every time.
Some participants got a message after round two. One version was an apology. The partner said it had screwed up, shouldn't have done it, and was sorry for taking so much.
The apology on its own made no significant difference. Not early in the recovery. Not by the end. Partners who apologized and then behaved well ended up in the same place as partners who said nothing and behaved well.
Read it twice. The words added nothing. The behavior did the work.
Three findings stood out to me:
- Consistent good behavior restored trust, with or without any message.
- A promise to change sped up the early recovery. Over time, the actions did the same job on their own.
- When the partner had lied as well as taken the money, trust never fully came back. Not with a promise, not with an apology, not after rounds of good behavior.
The researchers put it plainly. A promise signals you intend to change. After a run of observed behavior, the actions send the same message by themselves.
Not every apology is equal
Before you decide apologies are pointless, look at what the Wharton apology left out. It said sorry. It owned nothing specific. It offered nothing.
Roy Lewicki and his colleagues at Ohio State tested six parts of an apology on 755 people across two experiments:
- Expression of regret
- Explanation of what went wrong
- Acknowledgment of responsibility
- Declaration of repentance
- Offer of repair
- Request for forgiveness
The more parts an apology had, the better it landed. Two mattered most. First, acknowledgment of responsibility... saying plainly it was your fault. Second, an offer of repair. Lewicki's argument: talk is cheap, so commit to undoing the damage.
The weakest part? Asking for forgiveness. Lewicki called it the one to drop if you have to drop one.
Put the two studies side by side. An apology which says "I feel bad" does almost nothing. An apology which says "I did this, and here's how I'll fix it" opens the door. Then the fixing does the work.
Integrity breaks run deeper
Lewicki's team split the scenario. Half the participants read about a competence failure... a job candidate who didn't know the tax rules and filed a client's return wrong. The other half read about an integrity failure... the same candidate filed it wrong on purpose. Same apology. People accepted it less when the failure was about integrity.
Kim, Ferrin, Cooper and Dirks found the same split in 2004. Apologizing repaired trust better after competence failures than after integrity failures.
So be honest with yourself about which kind of break you made.
Missed a deadline because you misjudged the work? Competence. Own it, fix it, show the fix.
Told the team the reorg wasn't happening while you were planning it? Integrity. Now the Wharton data says you're climbing a hill with no top. You still climb it. You climb it for longer, and you stop expecting a finish line.
What "no one-and-done" looks like on a Tuesday
Grace's line sounds simple. The hard part is the calendar. Here's what I'd do.
Name it once, properly. Say what you did. Say it was your fault. Say what you'll do to fix it. Skip the speech about how bad you feel. Your team isn't there to manage your guilt. I wrote about the day I was wrong in front of my boss's boss. Owning it in the room was the start. It was never the whole repair.
Make one small promise, then keep it where people see it. The Wharton promise sped up early recovery. Pick something your team will watch you do within a week. "I'll send the budget numbers to everyone on Friday, not only to the leads." Then send them on Friday.
Keep the streak going after people stop watching. The first two weeks are easy. Everyone's paying attention, you included. Week six decides whether trust comes back. The repair happens in the boring rounds.
Don't ask for credit. The request for forgiveness ranked last in Lewicki's study. Asking "are we good now?" hands the work back to the person you let down. Let them decide when you're good.
Expect the scar. If you lied, the relationship won't snap back to where it was. Accept a different, slower trust. Don't punish people for being careful around you. You taught them to be careful.
Trust first, even after you've been burned
Grace's idea runs both ways. Leaders break trust. Team members break it too. Somebody hides a mistake, misses a commitment, covers for a friend.
The temptation is to go cold. Pull back their access. Check every piece of their work. Keep them at arm's length for a year. I understand the instinct. I've felt it.
Run the Wharton experiment in reverse. If you never pass the money again, the other person never gets a round to prove anything. Trust doesn't come back unless somebody extends a little of it first. Pass a small amount. Watch what comes back. Pass a bit more.
Going cold feels safe. It isn't. Every task now costs you a second check, and every person on the team learns mistakes get you frozen out. The lesson spreads fast. I wrote about what it does to innovation in Want Innovation? Stop Punishing the Mistake.
The apology was never the repair
In my own research, 99.5% of people said they'd had one or more types of bad boss. I don't believe most of those bosses set out to be bad. Plenty of them broke trust once, apologized once, and then wondered why the team never warmed back up.
They mistook the receipt for the repair.
A dry-stone wall goes back up one stone at a time. Nobody admires the first stone. Nobody notices the fortieth. The wall only shows up when enough of them sit in place.
Count your rounds
Think about the last time you let your team down. Not the time you'd like to talk about. The one you'd rather forget.
Now count the rounds since. How many times have you shown up, in a way they saw, and done exactly what you said you'd do?
If the honest answer is "I apologized," you haven't started yet.