You stay five minutes into one presentation dat going good. Da room stay with you, da argument stay landing, and you get answers ready for da questions you wen expect. Den you say one number out loud, and somebody in da second row get da real one on dea screen. Um forty, not sixty, and now everybody know. Get one small pause while da room figure out what you going do about um.
Nothing about da next hour stay decided yet. What happen next get decided by about sixty seconds, and dose sixty seconds get one script dat work. Fix um in one sentence, same volume, and keep going.
Rooms move on from one wrong number faster than anybody expect. What dey keep stay how da person wen handle um, and dat part stay entirely yours fo' write.
What I say right after I get one number wrong in front of everybody?
One sentence, in dis shape: da correction, den da right fact, den straight back to what you was saying. "I had dat wrong. Da number stay 40, not 60. So da second quarter stay da one fo' watch."
No preamble. No "sorry, I terrible with numbers." No story about where da wrong figure wen come from, which nobody wen ask for and which keep da error up on da screen for another thirty seconds. Da correction stay da shortest useful thing you can say, and da second um said, da room need you fo' be somewhere else. Dat's why da third clause matter as much as da first two. Moving on stay da signal all by itself. Um tell everybody da error was ordinary.
Volume matter mo' than people expect. Say um at da same level you wen use for da error. Dropping your voice stay da common instinct and um do da opposite of what you like, cuz one quiet correction read like one embarrassed one, and one embarrassed correction tell da room da mistake was one big deal.
Da same three parts work on one stage same as in one meeting, and dey work when somebody else catch um instead of you. If one person in da second row say da number stay forty, you no defend um and you no thank dem for one long time. "You right, um 40. Dat make da second quarter da one fo' watch." Six seconds, no drama, and da room now get two facts about you: da figure stay corrected, and you stay somebody who update in public with no fuss. Both of dose worth mo' than being right about da number would have been.
I supposed to apologize, or just fix um?
One short acknowledgment help and one long apology no help. Peter Kim and his colleagues wen study how trust get repaired and wen find dat when da failure stay about competence, getting something wrong instead of being dishonest, one apology repair trust bettah than one denial do.
So "I had dat wrong" stay doing real work. Um four words, um accept da error plain, and um make you da person who correct dea own record instead of da person who had to get corrected by somebody else. What none of dat research back up stay da apology tour: da follow-up email, da second apology at da end of da meeting, da mention of um three days later. Every repeat ask da room fo' look at da mistake again, and unlike da first one, none of dem add any information.
Da line stay easy fo' hold in practice. Acknowledge once, correct once, and never bring um up again unless somebody else do.
Why da correction gotta include da right answer?
Cuz one correction with no replacement leave da wrong fact just sitting dea. Irene Kan and her colleagues, studying how people update after dey get told something was false, wen find dat giving one alternative was da one thing dat consistently cut down da lingering pull of da wrong information, and dat pairing dat alternative with one direct correction wen work da best of all.
Dat's da whole case for da middle clause of da script. "Dat number was wrong" leave da room holding sixty and one question mark. "Da number stay 40, not 60" replace um. Saying da wrong figure and da right figure together stay what let people overwrite what dey already wen write down, which stay what you actually like, cuz half da room get sixty in dea notes.
Um also explain why "I going follow up on dat" stay weaker than um sound when you already know da answer. Follow up when you genuinely no get da figure. When you get um, say um now, in da same breath, while da room still holding da question.
What if da mistake was bigger than one number?
Den separate da fact from da recommendation out loud, and say which one wen change. "Da refund number was wrong. Um 40, not 60. Dat no change what I recommending, and here's why."
One wrong figure stay one competence slip and da room treat um like one. What make people nervous stay not knowing whether da conclusion still safe, so answer dat before dey gotta ask. If da conclusion do change, say so in da same sixty seconds and say what you would do instead. One person who revise one recommendation right dea on da spot, with one reason, stay demonstrating exactly da thing da room stay trying fo' figure out about dem.
Where da error stay about something other than accuracy, one missed commitment, something you said you wen do and you neva, dat's one different repair and um take longer than one sentence. Da same research on trust make dat distinction: da plain acknowledgment dat fix one competence error no stay da tool for da other kind.
What if I only notice one hour later?
Fix um in da same channel, at da earliest natural moment, and keep um to two sentences. "One correction from dis morning. Da refund number stay 40, not 60, and um no change da recommendation."
Same channel matter. If da error wen happen out loud in one room, one quiet message to one person leave da other eleven holding da wrong figure. If um wen happen in one document, fix da document and add one line saying you wen fix um. Do um before somebody else do, cuz da difference between "I had dat wrong" and "actually, dat was wrong" said by another person stay da difference between somebody who audit dea own work and somebody who need auditing.
What da room actually remember
Not da number. Almost nobody in dat room could tell you next week what da wrong figure was, and couple of dem neva even notice until you wen say something.
What dey keep stay one fifteen second impression of how one person behave when dey get caught out, and dat impression get used later, in rooms you not in, when somebody ask whether you can be trusted with something bigger. Handled right, one public correction stay one of da cheapest ways get fo' prove your numbers worth believing, cuz you just wen show da room you would tell dem if dey was not.
Get one version of dis dat compound in da other direction, and um worth naming so you can dodge um. Da person who never correct nothing in public no stay da person with no errors. Dey stay da person whose errors get found later by somebody else, usually at one worse moment and in front of mo' people. Correcting fast stay da cheaper of da two, every single time, and da cost drop as you do um mo', cuz one room dat wen see you do um twice stop treating um like one event at all.
One mo' thing, since dat's what people actually worry about afterward. Running da correction back in your head for da rest of da day do nothing for da room and take da edge off da next two hours of your work. Da sixty seconds stay da whole intervention. Once dey spent, da useful thing left fo' do stay check where da wrong number wen come from, fix dat, and get on with da day.
So use da sixty seconds. Say da correction, say da right fact, keep going, and let da rest of da meeting be da part people remember. Den get back to da work you was doing before, which was going good, and which still going good.
Sources
- PubMed, Removing the shadow of suspicion: the effects of apology versus denial for repairing competence- versus integrity-based trust violations
- Cognitive Research: Principles and Implications, Exploring factors that mitigate the continued influence of misinformation