The Summit Doesn't Pay

A coffee shop handed out ten-stamp cards: ten purchases, the eleventh free. Researchers collected 949 redeemed cards — roughly ten thousand coffee purchases — and looked at the intervals between them.

People sped up as they approached the reward. The gap between the first interval and the last was 0.7 days. This is not about loving coffee: customers given identical cards with no reward at the end slowed down instead.

What came next is more interesting. A hundred and ten people redeemed a first card and started a second. By the end of the first card they were buying coffee every 2.2 days. On the opening steps of the second, every 3.1.

Exactly the starting pace. As if the first card had never happened.

This is a follow-up to the piece on effort. There, effort broke a skill by loading it with an observer. Here effort works exactly as intended, all the way through, and zeroes out at the finish line.

The payout is for speed, not for arrival

The simplest explanation: the reward is issued not for having arrived but for closing the distance.

There is a formal version of this idea. Carver and Scheier proposed that affect tracks not the distance to a goal but its first derivative — the rate at which the gap shrinks. Moving faster than expected feels good. Slower feels bad. Standing still because you have already arrived is zero.

A caveat on the evidence, because this is an elegant model on a thin base. Direct tests over thirty-five years are almost nonexistent; in the one lab test using fabricated progress feedback, only the negative half of the hypothesis held and the positive half did not. The authors of the model themselves wrote in 2019 that they were aware of exactly one published study testing their predictions directly. So treat it as a frame, not an established fact.

What is measured better is the link between progress and well-being in general. A meta-analysis of 85 studies puts the correlation between progress toward one's goals and subjective well-being at around 0.39 for working adults. For comparison: which goals a person holds — their own or borrowed, intrinsic or extrinsic — yields correlations in the range of 0.15 to 0.24.

Which means "moving or stalled" explains more than "did you pick the right summit." Inconvenient for the entire find-your-purpose industry, but that is what the data says.

The bar moves with you

The second half of the mechanism: what counts as enough rises along with what you have achieved.

Stutzer ran the numbers on Swiss data, 4,462 observations. People were asked what income they considered the minimum they needed. A 10% rise in income lifted that figure by roughly 4.2%. By the Leyden estimates he cites, the shift in aspirations eats about 60% of the expected gain from rising income.

The author lists the caveats himself, and they matter: the data is cross-sectional, reverse causation is possible — unhappy people may simply name higher requirements — and he found no suitable instrumental variables to separate the two.

But the order of magnitude is clear. The bar does not sit still waiting for you to reach it. It travels with you, taking more than half.

Put the two mechanisms together. The reward is paid for speed, and at the finish line speed is zero. The bar has moved, so the distance to the new "enough" is non-zero again. The odometer resets, the distance reappears. Hence the next goal a week after the last one, and the sensation of running while standing still.

Two instruments

Now the part that matters most, and it is not the part usually told.

Achievement really does change something lasting. Just not the thing people are waiting for.

The best study on this is Swedish, built on registry data: 3,362 lottery winners surveyed 5 to 22 years after their win, with the analysis pre-registered. The result split across instruments.

Life satisfaction rose: 0.037 standard deviations per $100,000 after tax, p = 0.009. The effect does not fade with time — not after five years, not after twenty-two.

Happiness: 0.016, statistically indistinguishable from zero. Mental health: 0.013, also zero.

The two measures correlate with each other at 0.86, meaning they behave almost identically in surveys. Yet they respond to money differently, and the data rejects the hypothesis that the effects are equal.

The same gap shows up on different material. Sheldon and Lyubomirsky followed 481 people across three waves after a significant positive change in their lives. Over six weeks, thoughts about the event settled from 3.98 to 3.21, the positive emotions attached to it from 3.98 to 3.56, gratitude from 4.12 to 3.51. Effect sizes between 0.37 and 0.49.

Meanwhile overall well-being across all three waves: 4.83, then 4.84, then 4.95. It did not move at all.

The person is not deflating. What deflates is one specific feeling about one specific event — and it should deflate, it has no reason to persist. The baseline underneath was never moving in the first place.

Which gives the honest description of what happens at the summit: achievement moves your evaluation of your life, not how you feel. Two different instruments. You are sold the first and you watch the second, because the second is the one you actually feel. And when its needle does not twitch, it looks like nothing worked at all.

What gets misreported

The topic has accumulated a set of facts everyone knows and almost nobody has checked.

Lottery winners were no happier a year later. That is the 1978 study, and almost none of that sentence is in it. There was one measurement, not one a year later — a cross-section, nobody was surveyed twice. A before-the-win data point does not exist at all. There were 22 winners and 22 controls, with response rates of 52% and 41%. On present happiness the winners actually scored higher than controls, 4.00 against 3.82 — the difference simply did not reach significance at that sample size. Exactly one result in that study was significant: winners rated mundane pleasures lower, 3.33 against 3.82.

We now have Swedish data with a sample a hundred and fifty times larger and a pre-registered analysis. The 1978 paper is worth dismantling as an example of a study cited far beyond what it showed, not citing as proof.

People overestimate how happy an achievement will make them. Not in that form. A meta-analysis of 84 studies and 9,870 people found that the overestimation appears under one kind of question only: when someone is asked how they will feel "in general," without the event being mentioned. Ask about feelings about the event and the forecast is accurate, with an effect near zero.

What people predict badly is not the size of the joy. It is how quickly the event will stop occupying their mind. Two different errors, and the second one is far less flattering.

The "arrival fallacy." An elegant term from a 2007 popular book. A full-text search of the research databases returns a handful of hits for the phrase, all of them essays and commentary; not one empirical test. It appears in no diagnostic classification. As a label it works fine. As a hook for "studies show," it does not.

This is not damage

Before fixing it, it is worth confirming something is broken.

That same pace reset was tested in the lab — ten studies, 1,098 people. The post-reward dip is real: on the first step after the reward, activity drops significantly. On the second it is already marginal. After that it recovers, and the total effect of a temporary incentive comes out positive.

The mechanism, by the authors' own reading, is not undermined motivation. People simply want a break.

So the reset at the start of a new card is not psychological damage, and not evidence you were chasing the wrong thing. It is the release after tension. The trouble starts not when a dip follows an achievement, but when someone diagnoses that dip and buys treatment for it.

What it usually gets treated with

This is where I should offer a practice. The problem is that the practices do not hold up here, and the numbers are worth knowing before you pay.

Gratitude is the most studied. Compare it against doing nothing and there is an effect. Compare it against any other meaningful activity of the same volume and it disappears: d = −0.03 in the 2016 meta-analysis against a psychologically active control. A pre-registered 2025 meta-analysis covering 24,804 participants across 28 countries: overall effect 0.19, and against another positive activity, −0.15 and non-significant.

WOOP, otherwise known as mental contrasting with if-then plans, works: 24 effects, 15,907 people, g = 0.336. But the funnel is asymmetric, the publication-bias test comes in at p < 0.01, and after correction the effect drops to 0.242. On top of that, the do-it-yourself-from-instructions version is half as strong as the version run by a live human being.

Positive psychology as a whole: when the two most-cited meta-analyses were corrected for small-sample bias, the effect on well-being fell from 0.29 to 0.10, and the effect on depression from 0.31 to −0.03.

None of this means "it doesn't work." It means "it works about as well as any other meaningful activity you would spend the same time on." That is the honest price of admission.

What to do

Three things, and all three are about no longer expecting the summit to pay.

Count speed, not milestones. If the reward is issued for closing distance, the accounting should be about closing distance. Not "launched the product" but how many releases this quarter. Not "landed a client" but how many conversations this week. Milestones are rare by definition, and between them the counter reads zero — not because nothing is happening, but because the wrong thing is being measured.

Write the bar down beforehand, not after. Before you start, record what result you would consider sufficient. In writing, dated. When you get there, compare it with what feels sufficient now. The gap between those two numbers is the gain the moving bar ate. That is the only way to see it: from the inside the shift is imperceptible, and the new bar feels like the same one, the one that was always there.

Don't buy a feeling disguised as an evaluation. Achievement reliably moves how you assess your life as a whole, and barely moves how you feel on a Tuesday afternoon. That is not a defect in achievement. They are different instruments. If what you need is for Tuesday to get easier, the summit is the wrong tool, and no quantity of summits will substitute for it.

And here is the point of all this. The emptiness at the finish line is not a signal that you ran in the wrong direction. More often it is exactly what it looks like: the tension is released, the speed is zero, the bar has already moved ahead. A step or two later it comes back. The mistake was not arriving and feeling nothing. The mistake was expecting payment at the one place that does not issue it.

One thing here is nearly impossible to do alone: keeping the gap between the bar and the result visible.

The bar moves silently. Nobody has ever caught themselves thinking "half of this would have been plenty for me before" — because a new bar does not feel new, it feels obvious and permanent. Someone who in January wanted their first ten clients has forty by June and believes nothing is working. And they are not lying to themselves. They are honestly comparing forty against today's "enough," not against January's.

That is why the number has to be written down in advance and shown to someone who has no stake in defending your version of events. Only an external count and an external view put back what quietly moved.

That is what Strategic Momentum is built from: counting speed instead of milestones, criteria fixed in advance, regular review with someone on the outside. If you recognized yourself in the person who arrived, looked at the result, and could not work out why nothing changed — that is exactly the work we do there.