A decision you made three weeks ago is going badly. The numbers say so. The silence in the group chat says so. One teammate keeps re-asking a question you already answered, which is its own kind of answer.
And you happen to be the person who argued for that decision in a room full of people. So you keep the plug in and go hunting for reasons it might still work. This instinct has been measured in a laboratory, and it has a price tag attached to it.
Why 240 students kept funding something they knew was failing
In 1976, Barry Staw put 240 business school students through a role-playing exercise where they allocated research money to one of two company divisions, learned the division had performed poorly, and were then asked to allocate a second round.
The students who had personally made the first call poured significantly more money into the failing choice than the students who had inherited someone else's decision. Staw published it in Organizational Behavior and Human Performance under the title Knee-Deep in the Big Muddy, and the finding named something you have almost certainly felt: authorship, more than evidence, is what makes reversal hard.
Read that again, because it changes what the problem actually is. Reading the data was never the hard part. What makes it hard is that your name is on the thing being read. Everyone in that study saw identical results; only the ones who had authored the original call kept feeding it. Which means the fix is not more analysis. More analysis just gives a committed mind better material to build its defense out of.
The economist John Kenneth Galbraith put it more plainly in a 1971 essay collection: "Faced with the choice between changing one's mind and proving that there is no need to do so, almost everyone gets busy on the proof."
79 days that should have ended two careers
April 23, 1985, Lincoln Center. Roberto Goizueta and Donald Keough stood in front of the press and announced that Coca-Cola was retiring a 99-year-old formula. The decision was not reckless. It rested on roughly two years of work and taste tests with close to 200,000 consumers, and those consumers preferred the new version. Then the public revolted. Calls to the company hotline climbed to about 8,000 a day, and around 40,000 complaint letters arrived.
Seventy-nine days later, on July 11, the same two executives walked back to a podium and put the original formula back on shelves as Coca-Cola Classic. Keough did not blame the research or the market. He said the company had failed to understand how deeply customers felt about the product.
Now notice what did not happen. Neither man lost his job. Classic went straight back to leading the category. The reversal was fast, public, and specific about what had been missed, and it arrived with a product back on shelves the same week. Staw's students got punished for the slow version of this. Coca-Cola paid almost nothing for the fast one.
Reverse on evidence, hold on discomfort
Here’s how you can guide yourself on when to reverse and when to hold on.
Reverse when information exists now that you did not have when you decided. New numbers, a new constraint, a customer response nobody modeled. That is the Coca-Cola case, and it is the only case that earns you credibility.
Hold when the only thing that changed is your comfort. Slow early results, one loud critic, a bad week, a teammate who disagreed from day one. Those are conditions, and conditions move around on their own. Reversing on them teaches your team that the loudest objection wins, which costs you more authority over six months than any single wrong call ever will.
To tell the two apart, strip your authorship out of the question, since authorship is exactly what Staw showed distorts the call: "If this landed on your desk today with no history attached, would you still choose it?" If the honest answer is no, you already know.
Four lines that make a reversal sound like a decision
Say them in this order.
- Lead with the information. "Signups from the campus channel came in at 11 percent of what the model projected."
- Name your own call, in plain words, without softening. "That was my recommendation on March 4."
- State the new decision and the date it starts. "Starting Monday, that budget moves to referrals."
- Attach the trip wire for the new call. "If referrals are under 40 by the 20th, this changes again."
Open with an apology, and the room will rush to reassure you, and the meeting quietly turns into a conversation about your feelings. Start with the number, and it stays a conversation about the work.
Line 2 matters for the same reason: vague ownership ("we may have gotten ahead of ourselves") reads as an attempt to spread the cost, and people notice. Naming the date and the meeting where you made the call signals that you have already stopped defending it. Line 4 is the one people skip, and it is the one that turns you from someone who flip-flops into someone who runs decisions on evidence.
Write the exit condition before Friday
Pick the one decision currently running that has your name on it. Open the doc, the ticket, or the plan, and type one sentence at the very top:
"This is the wrong call if [specific measurable thing] is still true on [date]."
Then send that single line to the person most affected by the decision, with the subject "Trip-wire for the [X] call." Do it before Friday, while the decision is still going fine.
That sentence costs you nothing today and buys you the right to reverse later without a defense speech. Coca-Cola got to write its own reversal in 79 days. Wait long enough, and somebody else writes yours.
Until next time, Future Leader
