Why do I promise more than I can actually deliver?

Why Is a Promise the Cheapest Thing You Own at the Moment You Make It?

Because it is the only currency in the room that costs nothing to spend and settles instantly.

Consider what else is available to you in that half second. You could say no, which costs a visible disappointment right now. You could name the true number, which costs a smaller disappointment right now plus whatever the smaller number seems to say about you. You could ask for time to check, which costs a small awkward pause and makes you look less decisive than you would like. Or you could promise something larger than you can do, which costs precisely nothing in the room, produces immediate visible relief in another person, and settles the conversation on the spot.

Every behavior that becomes automatic has that profile. The reward is immediate, reliable and requires no cooperation from anybody. The cost is delayed, uncertain in size and lands partly on somebody else. A system that learns from consequences will pick the promise every time, and it will pick it faster each year, which is why fluency in this grows rather than fades.

There is a piece worth stating directly, because it does most of the work in the roots below. A promise whose outcome sits less inside your control is the weaker promise to make, and it is weaker in exact proportion to how much of the delivery depends on things you do not run. Most inflated promises fail this quietly. The date you named assumed a clear week you do not control, a client who responds when you need them, a body that does not get sick, and a version of your attention that shows up on demand. The promise sounded like a statement about your effort. Most of what it actually committed to was outside your effort entirely.

That single criterion, checked before the sentence rather than after, catches a surprising proportion of these before they happen. Not by making you promise less. By making you promise the part you actually run.

Why Does the Same Inflated Promise Come From Five Different Places?

Because a spoken commitment carries no information about what produced it, and five very different needs all resolve into the same date.

This is the limitation that makes advice about over-promising so unreliable in general. Someone watching you say Friday has exactly one observation available: the word Friday. They cannot see whether it came from a rough estimate, from a wish not to watch a face fall, from a self-description you are maintaining, from a genuine misread of next week, or from wanting the conversation to end. Neither can you, most of the time, because the production happened faster than introspection.

Only one of those five is a forecasting problem. That is the whole reason a general policy of being more realistic works so badly. It aims at the number, which is identical in all five cases, rather than at what the number was for, which differs in every one.

Notice too that the last column is available to you tonight without a conversation with anybody. Nothing there requires disclosure, apology, or a decision. It requires reading your own recent week honestly for about ten minutes.

What If You Meant Every Word of It?

Then the promise was an accurate report of your feeling and a poor report of your week, and both of those things can be true at once.

This is the root people are most relieved to find and most surprised by, because from the inside it does not feel like a pattern at all. It feels like caring about somebody. Somebody you like is in difficulty. You want to help. The wanting is genuine, unmixed, and the best thing about you. And the offer that comes out of it is sized by the wanting.

Koehler and Poon's 2006 finding is the research version of exactly this. Self-predictions get built from the strength of the current intention, with the situational factors that will actually determine the outcome weighted far too lightly. Your intention at that dinner table was extremely strong. It was also, as a forecast, almost information-free about a Tuesday in three weeks with a deadline in it.

There is a specific tell for this root and it is clean. The size of your promises tracks your affection for the person rather than the difficulty of the task. You promise more to friends than to clients, more at the end of a good conversation than at the start of one, more to somebody visibly struggling than to somebody managing fine. If you laid out your last dozen commitments and sorted them by how much you liked the person, the sizes would sort with them.

What helps here is smaller than it sounds and it does not require caring less. It requires putting a gap between the feeling and the number. The feeling is accurate and worth expressing immediately. The number is a separate object and should be produced somewhere the calendar is visible. Say the warm thing in the warm room, in full, with no hedging at all. I would really like to help with this. Then say the number tomorrow. Those two sentences have almost always been fused in your experience, and they come apart more easily than you would expect.

What If You Cannot Watch Their Face Change?

Then the promise is moving a cost from now to later, and later is where it becomes somebody else's problem as much as yours.

This root has a physical quality that people recognize instantly. There is a moment in a conversation where you can see a disappointment about to happen. It has not happened yet. You have perhaps a second in which you can prevent it, and preventing it is trivially easy, because all it takes is a bigger number. So you say the bigger number and you watch the disappointment not arrive, and something in you settles.

What actually happened is a transfer. The disappointment was not removed, it was scheduled. It will now arrive in three weeks, larger, attached to a failure rather than to a limit, and it will arrive in a week where somebody else has already planned around your number. A disappointment absorbed now is one conversation. A disappointment delivered late is one conversation plus a wasted plan plus a revision to how reliable you are considered to be.

The tell is that the dread is wildly out of proportion. A four-hour task should not produce a Sunday evening of low-grade avoidance. When it does, what you are dreading has very little to do with four hours of work. What you are dreading is the moment where they find out.

There is something worth knowing about what actually happens on the other side, because the imagined version is much worse than the real one. People calibrate reliability against what was said, not against some absolute standard of helpfulness. A modest promise delivered on time reads as competence. An ambitious one delivered at sixty percent reads as a pattern. Almost everybody who tries naming a smaller number for the first time reports the same thing afterward, which is that nothing happened. The face did not fall. They said fine, and moved on, and the entire cost you had been paying to avoid was imaginary.

If the difficulty is specifically that a refusal or a limit feels dangerous rather than merely uncomfortable, the avoidance of the conversation itself is the more useful subject, and if the whole pattern only ever appears in response to what other people want, people pleasing is the wider frame.

What If You Were Forecasting for a Version of Yourself Who Never Turns Up?

Then you have the one root that is genuinely a math problem, and it is the most tractable of the five.

Here the promise is not buying anything at all. Nobody's face was involved. You made an estimate, you believed it completely, and it was wrong. The version of you who does the work in your forecast is a person with a clear week, full attention, no interruptions and no competing obligations. That person exists occasionally. They do not exist on the specific Tuesday you committed to.

Zauberman and Lynch reported in 2005 that people expect to have considerably more spare time in the future than spare money, and commit time far more freely as a result. Next month reads as empty in a way that next month never turns out to be. That single asymmetry explains a great deal of ordinary over-committing, and it explains why a date six weeks out attracts a promise that the identical task next Tuesday would be refused.

Bazerman and Tenbrunsel described a related split between the self that decides and the self that acts, which are answering different questions under different conditions and rarely consult each other. The forecasting self is calm, has an overview, and is imagining. The delivering self is tired, has four other things live, and is inside the week.

The tell is that this root does not discriminate. It happens with strangers, with people you dislike, and with tasks you have no feeling about at all. The scope you promise is usually about right. The date is what is wrong, consistently, in the same direction.

The fix is unglamorous and it works. Write the estimate down before you say it, break the task into its actual parts rather than treating it as one lump, and add the parts you never count, which are the waiting, the review, the thing that comes back needing changes, and the first hour of getting back into it. Then look at the total and say that number out loud instead of the one in your head. If your commitments are correctly sized and simply arrive late every single time, the full arithmetic and the segment-by-segment method live on the lateness page, and that is the better tool for this root than anything here.

Why Does the Number Get Defended Instead of Corrected?

Because the honest revision gets more expensive every day it is postponed, while the reason to postpone stays exactly the same.

Watch the arithmetic of a slipping promise. On day two you know the date is wrong. Correcting it on day two costs one message and a small amount of embarrassment. On day nine, correcting it costs one message, the same embarrassment, plus an explanation of why you waited a week. On day sixteen it costs all of that plus the fact that they have now planned around a date you knew was wrong for two weeks. The price of honesty rises steadily and the price of silence stays at zero right up until the deadline, at which point it becomes very large all at once.

Staw's work on escalation of commitment describes the same shape in organizational decisions, and Brockner's 1992 review found it robust across settings. Once a course has been publicly chosen, evidence that it is going badly tends to produce further commitment rather than revision. Applied to a Friday you named at a kitchen table, this predicts precisely what people report: two weeks of intending to send the message, and then a delivery at sixty percent on the day itself with an apology attached.

There is a second mechanism stacked on the first. The revision message feels like it requires an explanation, and explanations invite discussion. Discussion is exactly what you did not want, which was often the reason the promise got made in the first place. So the message keeps not getting written, because in your head it is four paragraphs long.

It does not need to be. The version that works is one line with a number in it. The date has moved to the twenty-second. The version that fails arrives late, carries several sentences of reasoning and no new date at all, which leaves the other person holding uncertainty rather than information. Give them the number and let them plan. And if the apology has become the main event, arriving in four sentences every time, the apology itself has its own mechanics and its own page.

Why Has Deciding to Be More Realistic Never Worked?

Because it aims at the wrong layer, at the wrong speed, and it treats a five-cause pattern as though it had one cause.

Take the speed first. The inflation happens inside the sentence, faster than deliberation arrives. A general resolution to be realistic lives in deliberation. It will be present and effective in every conversation where you were never going to over-promise, and it will be absent in the warm ones, which are the only ones that matter.

Then there is the aim. Deciding to be realistic addresses the number. The number is identical across all five roots, so the resolution is fired at the one part of the pattern that carries no information. For the forecasting root it happens to be roughly the right target and it still fails on speed. For the other four it is aimed at the wrong object entirely.

There is a third reason and it is the one that catches thoughtful people. Insight into which root you have changes nothing on its own. The inside is only going to stay on the inside unless you make it known and public on the outside and back the idea up with action. And the required action is different in each case. For the warmth root it is separating the feeling from the number by a day. For the disappointment root it is one smaller number said once to somebody safe. For the identity root it is a single piece of evidence about what a smaller number actually costs. For the forecasting root it is writing estimates down before speaking them. For the exit root it is one rehearsed sentence that closes a conversation without committing to anything. There is no substitute for the right action, and the right action differs in every one of the five, which is the entire reason for spending ten minutes finding out which is yours. The general problem of understanding something and continuing to do it anyway is covered separately.

Expect something specific when you start, because it catches people out and gets misread as failure. When you stop making the inflated promise, the warm moment will feel flat. Something that was reliably producing a small good feeling has been removed, and nothing has replaced it yet. It is actually not that hard once you commit to it, but it does feel like there is something missing in the meantime. That gap is the ordinary shape of removing anything that was quietly doing a job, and it closes.

What Would You Have to Give Up to Say the True Number?

More than you would expect, and naming it in advance is what stops the change collapsing in week three.

This is the question most advice skips, and skipping it is why the resolutions fail. Something is being given up, it is worth having, and pretending otherwise makes the loss feel like evidence that you have chosen wrong.

You give up the instant relief. The half second where the face changes back is genuinely pleasant, it arrives reliably, and nothing else in the conversation supplies it.

You give up a version of yourself. If you have been the person who can, the true number retires part of that description, and there is a real adjustment in being someone who names limits out loud rather than someone for whom limits do not seem to apply.

You give up the exit. Some conversations will now run longer, because the sentence that used to end them is no longer available.

And you give up a certain kind of hope. An ambitious number carries a private wish that this time you will be the version of yourself who does it in five weeks. Naming eleven weeks means giving up on meeting that person on this particular occasion.

What arrives instead takes longer to notice and is worth considerably more. Your dates start being used as inputs by other people. Your Sunday evenings stop containing three items of low-grade dread. And after about six months of promising numbers you believed, you find out for the first time what your actual throughput is, which is information you have never had and cannot get any other way.

What Does a Promise You Can Keep Sound Like on an Ordinary Tuesday?

Undramatic, slightly shorter than you are used to, and with a number in it rather than a feeling.

If the middle column reads as out of reach tonight, hold off on any verdict about yourself. The left column was built by a real life under real conditions, every entry in it was solving something at the time, and most of what it was solving was small and reasonable.

One caution about the second row, because people overshoot it. Delaying the number does not mean withholding the warmth. The warmth should be said immediately and in full. What moves to tomorrow is the date, and only the date.

How Long Does This Actually Take to Shift?

Faster than the guilt version suggests and slower than the resolution version does, and it runs at five different speeds.

Noticing arrives quickly. Ten days with a written list of every commitment you make and most people can name their root, name the two or three settings it clusters in, and describe the exact point in a sentence where the size goes up. That stage is uncomfortable in a specific way, because you will catch yourself doing it live and say the bigger number anyway. Seeing it counts. Everything after it depends on it.

The exit root moves fastest, often inside a single conversation, because it needs one rehearsed sentence and nothing else. The forecasting root moves within a month once you start writing estimates before speaking them, and it moves on method rather than on effort, so the work is tedious rather than hard. The warmth root takes a few weeks of deliberately splitting the feeling from the number, and it feels artificial for about five attempts. The disappointment root needs one piece of real evidence, which means one smaller number said to one safe person, and it can move in a day once that evidence exists. The identity root takes longest, because the smaller number has to stop meaning something about you, and that changes slowly and only through instances.

Expect the count to go up before it goes down. Around day four the list gets longer rather than shorter, which is almost always a jump in noticing rather than in doing. That is the ordinary shape of watching a behavior rather than a sign that anything has gone wrong.

What Is Worth Doing Before Your Next Conversation?

One prepared sentence, a written list, and a single smaller number, and the first four days ask you to change nothing at all.

1. Write every commitment down for four days. Every date, scope and vague large yes, including the ones you make to yourself. Do not change any of them. You are collecting a baseline, and a baseline from a normal week is worth more than a good week.
2. Mark each one with the number you were holding. Next to what you said, write what you privately thought before you said it. The gap between those two columns is the actual measurement, and most people have never once written it down.
3. Run the shrink test on the three most recent. Build the smaller version in your head and note whether logistics or dread comes up. This one question separates a forecasting problem from everything else, and it takes about a minute each.
4. Sort by affection and by room. Does size track how much you like the person, or does it track being assessed? Those two answers point at separate causes, and the tests are independent of each other.
5. Prepare one delaying sentence and say it out loud alone. Let me look at my week and come back to you today. It has to be rehearsed, because the moment it is needed contains no spare thinking.
6. Say one smaller number to one safe person. Chosen for being low-stakes rather than important. Then watch what happens, which in almost every case is nothing.
7. Send one revision you have been postponing. One line, with a new date in it, no explanation. Notice how much less it costs than the version you had been drafting in your head for two weeks.

That sequence costs almost nothing and produces something no article can hand you, which is a measurement of your own gap in your own handwriting. Then put whatever you decide into your real calendar and hold the calendar, because a decision that lives only in your head has no way of surviving a Wednesday.

If what landed hardest was the shrink test, and specifically the discovery that some of your promises will not go smaller, our Sacred Valley seminar works at that layer rather than at the level of the number. What happens there is that we find what the promise was buying, since that is the part no estimating method reaches, and then build something that does the same job without a Friday attached to it. I cannot promise you that five days will make you a person who never over-commits again. I can certainly set you in the right direction, and I would rather say that than a better sentence I could not stand behind. Everybody is screened beforehand, which keeps the group small enough that somebody who has spent fifteen years saying larger numbers than they meant cannot quietly keep doing it inside the room. The weeks afterward are part of the design rather than something left to you, because the real work only starts when you come home to the same inbox and the same person waiting on the same Friday. One limitation worth stating plainly: exact dates, the price and the final cap on group size are still being settled, so if you are comparing programs on those numbers this week, ours cannot yet be compared. Ask every program you are weighing the same questions, including that one, and decide from your own reading of what comes back. The four-day list costs you nothing and is worth having whatever you decide afterward.