Life GPS Try the beta

You can win the metric and lose what it was for

Most of us are still running on instructions written for a child.

Grow. Learn faster. Get more. Go higher. Next level.

Those instructions are useful. The mistake is assuming they never need re-examining.

Some things should keep growing. Some become more valuable to maintain. Some need another route entirely. Some are worth releasing, so the effort can go where it still does something.

Nobody teaches us when to change the operation. So more keeps running by default — and everyone around you is running beside you, doing the same, at speed.

We taught people how to win the number.

We never taught them to check what the number was for.

Money. Grades. Followers. Weight. Titles. Revenue. Performance reviews. Every one of those can be a real goal, chosen deliberately, worth years of someone's life. Every one of them can also be nothing but a proxy — a visible, countable stand-in for something that was never visible or countable in the first place.

The question almost nobody is asked, at any point, by anyone:

What is this supposed to make possible?

Medicine has a name for this error

The name does not come from psychology. It comes from clinical research, where confusing a measurable stand-in with the actual outcome has killed people.

A surrogate endpoint is something you can measure quickly — a lab value, a rhythm on a monitor — used in place of a direct measure of how a patient feels, functions or survives. Regulators draw that line explicitly, and they still warn that even a validated surrogate can give a misleading picture of overall benefit and risk.

In the late 1980s, the Cardiac Arrhythmia Suppression Trial tested drugs against the irregular heartbeats believed to predict sudden death. They successfully suppressed the ventricular arrhythmias the trial was targeting. But success on that intermediate measure did not translate into clinical benefit: the trial was stopped early because mortality was higher in the treatment group than in the placebo group.

The number went the right way. The people did not.

We are not claiming a human life behaves like a clinical trial. We are pointing at the shape of the error, because it is the same shape:

The number can be right. The life can still be wrong.

A failing metric can warn you. A successful proxy can lie for years.

That is the asymmetry the whole argument turns on.

A metric that is failing can contradict itself — it stops moving, it goes the wrong way, something visibly does not add up, and that can eventually force a question. A metric that keeps improving never has to.

Income climbing. Followers climbing. Weight falling. Company growing. Grades good. Every quarter the instrument reports the same thing: working, working, working, working.

Meanwhile the thing the number was chosen to deliver may be deteriorating, and nothing in the system is watching it, because nobody built an instrument for it. The blind spot is not created by failure. It is created by success on the wrong measure.

FOMO does something more specific than make you hurry

The original research defines fear of missing out as the apprehension that others are having rewarding experiences you are absent from, and the desire to stay continuously connected to what they are doing. Later work connects it to social comparison.

But there is a mechanical detail inside it that matters more than the anxiety:

You see the result before you know the reason.

You can see someone's million. You cannot see what they wanted it for, what it cost them, or whether it delivered. You can see the title, not the reason the title mattered to that person, or whether it still does. Comparison transmits the visible part of someone else's life and strips out the part that made it make sense.

That is our claim rather than a finding: the research establishes the fear and the comparison, not this particular route. But the route follows from them. It creates a way for somebody else's proxy to become your goal before you have examined why you want it.

And it does not stop with you. Once you adopt the number, your result becomes the next visible signal for somebody else, who sees it without its reason, exactly as you did.

See → copy → optimise → repeat.

Which reframes what FOMO is doing. It is not only spreading anxiety about what you might be missing. It is helping metrics reproduce.

That also explains the thing that looks like mass irrationality from the outside — everybody running the same way at once. A culture does not have to tell people what to want. It only has to make the same outcomes visible enough. People do the rest.

The pieces of this are known separately — mimetic desire, goal contagion, the distinction between intrinsic and extrinsic aspirations, proxy failure in systems. We have not found them assembled into this particular mechanism, and we would genuinely like to be shown where they have been.

It travels through very different cultures

The tempting version of this argument is that some societies escaped it — the slower ones, the older ones, the ones with a different relationship to time. Start with the hardest case for that idea.

Bali. A place the world has spent two decades associating with slowness, spirituality and wellness. A 2025 study of school students in Denpasar put 40.7% at a moderate level of FOMO. A 2026 study looked at its role in Black Friday and flash-sale buying on the island. Another examined it inside wellness tourism in Ubud — the industry that sells escape from exactly this. Separate work on Gen Z TikTok users in Denpasar found that content personalisation and digital lifestyle were associated with higher FOMO, with social proof strengthening the effect.

Thailand. A Buddhist-majority society, with its own 2026 study of Gen Z, and national figures in which 31.2% of Thai social-media users named staying on top of trends and events — avoiding missing out — as a reason they use social media. None of that research tested religion as a cause, and neither do we. It simply does not disappear there.

And the cleanest evidence is a single study across eight countries using one measure, which is worth more than a pile of percentages from different scales. Mean FOMO was non-zero in every sampled culture — roughly from Singapore at the low end to China at the high end, with Indonesia, Thailand, Vietnam, Malaysia, the Philippines and the United States in between. The level varied. FOMO was measurable in every sampled country.

Two more points that keep this honest. Comparisons between countries refuse to fall into a tidy pattern — one study found higher general FOMO in Poland but higher workplace FOMO in China; a European sample scored above an Arab one in one comparison, while a Chinese sample scored above a US one in another. No simple collectivist-versus-individualist story explains the pattern. And a meta-analysis of thirty-three samples covering more than twenty-one thousand people found that geographic region did not moderate the link between FOMO and social-network use.

These are also, in several cases, convenience samples of students or users rather than national estimates, and we are not going to pretend otherwise.

What the evidence supports: FOMO appears across very different cultural environments, while its level and the contexts in which it shows up vary considerably.

Our hypothesis goes one step further: culture helps supply the cargo. It shapes which outcomes become visible, socially valuable, and therefore easiest for this mechanism to transmit. In one place what travels is a trend, a concert, a place to be seen. In another a grade, a university, the speed of a career. In another the standing of a family. Different cargo, same freight system.

FOMO may be human. The metric it carries is cultural.

Which gives the shape of the whole thing in three lines: culture decides what becomes visible. FOMO helps it travel. And the only question left is whether it should become yours.

And there is a reason the number travels so easily:

What is easiest to compare is what is easiest to count.

Love, calm, the feeling of being in the right place — these are visible, but they do not line up in a ranking. Income, weight, followers, job titles and grades do. Platforms did not invent comparison. They made it continuous, scalable and increasingly quantifiable — and the reason is far harder to transmit alongside it.

Choosing the number is not weakness. It is an information problem.

It is tempting to read all of this as a failure of character — people are shallow, people compare, people chase. That is both unkind and wrong.

The number is available today. The reason can only be verified later, sometimes decades later, and sometimes never. Faced with one signal you can act on now and one you cannot check until it is too late, choosing the measurable one is not stupidity. It may be the clearest feedback available.

The problem is not that people optimise. It is that one instrument can become dominant enough to be mistaken for success itself.

Where the instrument is handed over

For many people, a grade is one of the first formal numbers by which their performance is repeatedly judged.

Here it is worth separating what is established from what we are proposing, because the difference matters.

What the research supports is narrower than the headline: across a meta-analysis of sixty-eight studies and nearly forty-eight thousand students, perceived classroom goal structures are associated with the achievement goals students adopt. Environments that emphasise mastery and understanding go with one motivational pattern; environments that emphasise grades, competition and demonstrating ability go with another. Association, in context, over years.

Our claim goes one step further. A child in those rooms is not only learning mathematics, language and history. For thousands of hours they are also learning which signals count as evidence that things are going well.

And the hypothesis we would defend: the lesson outlives the lesson plan. The number changes — grade, salary, valuation, follower count — and the habit of looking to an external number for evidence may not.

Education does not only transmit knowledge. It helps establish what a person learns to recognise as evidence of success.

That is a heavier claim about schooling than the usual complaints, and it is the reason the education work exists.

And the systems around you are not neutral about your goal

Not every system that helps you pursue something is rewarded when the underlying thing is achieved.

An advertising-funded platform earns from continued attention. A subscription earns from renewal. A marketplace earns from transactions. A course earns from the next course. None of that requires bad intent, and sometimes the incentives line up perfectly with the person — a product that genuinely delivers gets recommended, keeps its customers, and wins.

The point is narrower: alignment is not guaranteed by the structure, so it has to be checked.

Which is the exact mirror of the argument we make elsewhere on this site:

A system can successfully optimise its metric while failing your outcome. So can you.

Sometimes the million is the goal

This is the moment where most writing on this subject goes wrong and starts telling people to want less.

The data does not support the comfortable story. There is no simple universal income ceiling beyond which money stops being associated with emotional well-being. In a large experience-sampling study of employed US adults, an adversarial collaboration between researchers who had publicly disagreed found that emotional well-being continued to rise with log household income for most participants across the observed range — with flattening concentrated among the least happy fifth, and the exact shape at high incomes still debated. Money is not a trick.

A separate body of work makes a different and more useful distinction. The US consumer financial regulator defines financial well-being not by income or net worth alone, but by financial security and freedom of choice, present and future — and notes that the same income can correspond to very different levels of it.

Which gives the general form of the problem: for many goals, the amount is not the same thing as the function the amount is expected to perform.

So the question is not the coaching-seminar version, do you really need a million? It is:

What exactly must the million make possible?

Time. Safety for your family. The right to walk away. A house. Care for a parent. Or the game itself — some people want to build the largest thing they can build, and that is a legitimate answer, not a symptom.

Every answer is allowed. The only thing not allowed is never having asked.

If after the question you still want the million — good. Now it is a chosen target rather than an inherited assumption.

The second way to lose, with the right goal

Suppose the goal survives the question. You examined it, and you want it.

You can still lose, because the method can destroy the purpose.

Someone builds the company they genuinely wanted to build, and arrives without the health to run it, without the relationships it was supposed to protect, or without the capacity to enjoy a single thing it bought. The goal was real. The strategy failed anyway.

This is not a metaphor about balance. The WHO and ILO estimate that working 55 hours or more per week is associated with increased risk of stroke and ischaemic heart disease. Long hours are sometimes the right trade — before a launch, in a crisis, for a season. The failure is not intensity. The failure is a cost that falls outside the horizon of the decision.

Elite sport makes its optimisation target unusually explicit: competitive performance. Outside such explicit systems, the target and its cost are far less clearly separated.

Reaching the destination is not enough if you destroy the person who was supposed to live there.

Growth is one operation, not the only one

There is a popular version of this argument that says: the brain finishes developing around twenty-five, so after that, stop chasing and start preserving.

That version is wrong, and it is worth saying so. Brain maturation does continue into the mid-to-late twenties, but different cognitive abilities peak at wildly different ages — some in youth, some in the thirties, some after forty. In sport, peak performance depends entirely on the demand, ranging from the early twenties to nearly forty. There is no single peak and no age at which a person switches modes.

What lifespan research does support is subtler and more useful: growth is not the only way to adapt. Maintenance, loss prevention, selection, compensation, and disengaging from an unreachable goal in order to redirect effort are all recognised strategies, and letting go of a goal that cannot be reached can be adaptive rather than a defeat.

So the useful distinction is not age. It is which operation a particular capability needs right now:

Grow what you have decided to grow. Maintain what is expensive or irreversible to lose. Adapt — keep the function by another route when the old route closes. Release — put the effort somewhere it can still do something.

All four can be true at once, in different parts of one life.

What is actually worth protecting

The WHO does not define healthy ageing simply as the absence of disease. It centres functional ability: the capabilities that let people be and do what they have reason to value, shaped jointly by a person's intrinsic capacity and the environment they live in.

That is the right object, and it is not a score. The line that follows is ours, not theirs:

Do not preserve the peak. Preserve the ability to use what the peak was for.

The final asset was never the number. It is the capacity to live the thing the number was meant to unlock.

Why nobody stops to check

Two mechanisms, neither of which is a character flaw.

Stopping inside a race reads as losing it. The check itself looks like the thing you were taught to avoid.

And the sunk-cost effect is real and well documented: the more money, effort and time already invested, the harder it becomes to change course. The longer you run, the more expensive it becomes to ask whether you chose the race.

Put those together with a proxy that keeps reporting success, and you get a person doing everything right by every instrument available to them, for years, without rechecking whether the direction still serves what they chose it for.

You cannot answer this once

Here is the honest objection, and it is a good one.

Research on affective forecasting shows that people often mispredict the intensity or the duration of their future emotional reactions. If that is so, how is anyone supposed to name in advance what the million must make possible?

They are not — not once, not in advance, not by thinking harder.

Which is exactly why this cannot be a moment of introspection. A forecast is a hypothesis, and hypotheses are updated by reality. It has to be a loop: you name what you think it is for, you act, reality answers, and the answer changes the goal. The point is not to guess correctly at the start. The point is to keep the question alive long enough for reality to correct it — while there is still time to use the correction.

And part of the answer is not available from inside your own head.

Research gives distributions. Lived experience gives context. Neither replaces the other. Someone who has already walked the path cannot tell you what is true in general — they can tell you what happened to them: what it gave, what it cost, and what they would notice earlier if they ran it again. That is a second kind of knowledge, not a better one, and it is missing from aggregate data entirely. Which is why the network of people matters here more than any amount of advice.

Two questions

If you take one thing:

What must this make possible?

How will I know when that job is done?

The second one is often missing from goal-setting entirely, and it is where enough actually lives. Enough is not a quantity and not modesty:

Enough is the point at which a means has done the job you chose it to do.

If the million was for security, enough is defined by whether you are secure — not by somebody else's balance. If the million was the game, enough is somewhere else entirely, and that is fine. Nobody else gets to set it, including us.

FOMO has a name. The check before the chase does not.

There is a word for the pull. There is no widely shared word for the check — people do reflect, clarify values and adjust goals, but none of that arrived with a label that spreads the way FOMO did.

So we gave it one, and we are deliberately not dressing it up as a scientific model — there are already models with this name, about other things. It is a practice, and it has two questions in it.

FOMO asks: what am I missing? SAFE asks: what is this for?

It is not the joy of missing out. That is a different move — choosing not to participate. SAFE is not about participating less. It is about keeping the metric connected to the reason you chose it, while you run.

FOMO spreads the result. SAFE restores the reason.

And because it is a thing you do rather than a thing you have:

Before you chase it, SAFE it.

Which is also the answer to the accusation this article invites. SAFE is not an argument against ambition. It is a way of stopping ambition from becoming inherited automation.

The last question arrives either way

FOMO looks forward: what am I missing?

Regret looks backward: what did I miss? The research finds an asymmetry over time — actions can sting more at first, while the things not done are more represented among long-term regrets.

They are not the same psychological construct. But they rhyme across a life: one asks what might be missed, the other asks what was.

A system built around a person should not wait for the second question to discover that the first one was pointing in the wrong direction.

None of this is permission to try less.

Checking the direction makes the effort more expensive, not cheaper — because afterwards you have to actually run.

Sources

Proxy and surrogate endpoints

FOMO, comparison and transmission

Across cultures

What environments teach

Money and what it is for

Cost, capability and modes

Why nobody stops

Published 2026-09-18 · Updated 2026-09-18 · Every claim on this page carries a status