Someone tells you cold showers changed their life. Someone else swears by the 5 a.m. wake-up, or the elimination diet, or quitting social media, or the gratitude journal. And the striking thing is that they're not lying. It genuinely did change their life. You can hear it.
So you try it. And you get about nine days in before it quietly stops, and the conclusion available to you is the one that's always available: other people can do this and I can't.
That conclusion is almost never correct, and this post is about why. Not "most advice is bad" — plenty of it is excellent. The problem is structural, and it sits in the space between the advice and you: advice is an answer, and you never got told the question.
Advice is the second half of a conversation
Every piece of advice that works is a solution to a specific diagnosis. Cold showers changed that person's life because of something particular about their situation — a nervous system stuck in low arousal, a morning with no structure, a self-concept that needed a daily piece of evidence about doing hard things.
That diagnosis is the load-bearing half. And it's the half that doesn't travel. What travels is the prescription, because the prescription is concrete, repeatable, and easy to say. Nobody opens with "I had an unusually unstructured morning and a low-grade belief that I couldn't tolerate discomfort, so what worked for me was…" They open with cold showers.
So the advice arrives stripped of the thing that made it work. You then apply the prescription to your own unexamined situation and, if your diagnosis differs, you get nothing — not because the advice is bad, and not because you're weak, but because you took someone else's medicine.
This is the whole reason this blog is organized around a dashboard rather than a ladder, and around diagnosis before prescription. Not because tiers are magic, but because you cannot evaluate a prescription without one.
Three structural reasons advice arrives miscalibrated
It's worth being specific about the mechanisms, because "advice is personal" is a bumper sticker and the actual dynamics are more interesting.
Survivorship in who gives advice. The people who become advice-givers are, almost by definition, the people it worked for. Someone tries the 5 a.m. routine, thrives, and starts telling people. Someone else tries it, is destroyed by it, and goes quietly back to sleeping until seven — and does not write a book about the time a routine didn't suit them. The advice you encounter has been filtered by outcome before it reaches you, so the population of available advice systematically overstates how well any of it generalizes.
Base rates you can't see. Suppose an intervention genuinely helps thirty percent of people. From inside a testimonial, that is completely indistinguishable from one that helps everyone — because you're hearing from a member of the thirty. And you have no access to the denominator. This is also why "it worked for thousands of people" is a much weaker claim than it sounds: thousands of successes is consistent with a low success rate and a large audience.
The n=1 problem, which is worse than it looks. Personal testimony is simultaneously the most persuasive form of evidence and among the least generalizable. It's persuasive because it's vivid and specific and comes with a face. It's unreliable because a single person's experience confounds the intervention with everything else about their life — and, less obviously, because the person reporting it is not a reliable narrator of why it worked. Research on self-knowledge is fairly humbling here: people are limited at identifying the actual causes of their own states, and we reach for the culturally available explanation over the true one (Wilson & Dunn, 2004). So the advice-giver may be sincerely, confidently wrong about which part of what they did was the active ingredient.
That last one deserves a beat. It means even honest testimony from someone the intervention genuinely helped can misidentify the mechanism. They changed four things and credited the memorable one.
The predictable failure
Here's the sequence this produces, which we'd guess is familiar.
You try the popular thing. It doesn't take. You try a different popular thing. That doesn't take either. Somewhere around the fourth attempt, the explanation stops being about the interventions and starts being about you — and now you're carrying a general belief about your own capacity that was assembled entirely out of mismatched prescriptions.
The belief is expensive because it generalizes. It doesn't just mean "the 5 a.m. thing isn't for me." It means I don't follow through, which is a claim about your character that will be applied to the next thing before you've started it.
And it's built on a mistake in the unit of evaluation. Each failed attempt was evidence about a match between one prescription and one unexamined situation. None of them was evidence about you.
Diagnose first, then prescribe
The alternative isn't complicated. It's just the reverse of the usual order.
Diagnose first. Find out which of your needs is actually thin before choosing an intervention — the diagnostic protocol is our version, four probes and about twenty minutes. Then pick advice aimed at that.
The gain isn't only that you pick better. It's that failure becomes informative again. If you've diagnosed a belonging deficit and a specific intervention doesn't move it, that's data about the intervention, and you try another one aimed at the same target. Without a diagnosis, every failure lands on your character, because there's nothing else for it to land on.
Notice this is what a good coach does. The evidence says coaching works — Theeboom and colleagues pooled the available studies and found positive effects on performance, wellbeing, coping, and goal-directed self-regulation in workplace settings (2014). It doesn't say why, and it didn't test diagnosis-first against prescription-first. That the diagnosis is the active ingredient is our claim, not theirs — but coaching is precisely the format where somebody spends the early sessions on your situation before recommending anything. The advice is comparatively cheap; a good coach could give you generic advice in ten minutes and it would be worth about what it cost.
There's a quieter benefit to diagnosing first, and it may be the more valuable one. It changes what you do with advice you decide not to take. Without a diagnosis, declining advice feels like avoidance — you're the person who was told what to do and didn't. With one, declining is a decision you can defend: this intervention is aimed at a need that isn't currently thin for me. That's the difference between discipline and discernment, and only one of them is available to someone who hasn't looked.
There's a related point worth one sentence. Advice adopted from a testimonial is, structurally, someone else's goal — which is a reasonable first guess at why it evaporates in week two.
How to read any advice you encounter
A practical filter. Four questions, usable on anything — including this post.
1. What would have to be true about me for this to work? The single most useful question, and it forces the missing half back into view. Cold showers: probably that you're under-stimulated in the morning and short on evidence that you can do hard things. If neither is true of you, you've learned something in ten seconds.
2. Does it name who it's not for? Advice that specifies its own preconditions and exclusions is advice that's been thought about. Advice presented as universal is either untested or overclaimed. This is the strongest single signal of quality, and it's rare, because caveats read as weakness in a marketplace that rewards confidence.
3. What's the mechanism? Not "does it work" but "what is it supposed to be doing?" If you can state the mechanism, you can check whether the mechanism applies to you — and you can often find a substitute that does the same job in a form you'll actually sustain.
4. What's the cost if it's wrong for me? Some advice is cheap to test — two weeks of earlier bedtimes costs almost nothing. Some is expensive and hard to reverse. Calibrate your scrutiny to the cost, not to the confidence of the person giving it.
None of these requires you to become a skeptic about everything. They just re-insert the diagnostic step that the format of advice-giving strips out.
When diagnosis is the wrong move
The argument has a limit, and it would be a poor post that didn't name it.
Some interventions are so cheap to test that testing beats diagnosing. Going to bed an hour earlier for two weeks costs almost nothing and answers its own question faster than any amount of self-examination. If the test is cheap, reversible, and quick to read, run it — the diagnostic step is overhead you don't need.
And there's a failure mode where diagnosis becomes the activity. Working out which tier is thin is more comfortable than doing anything about it: it feels like progress, it produces insight, and it can be repeated indefinitely. If you've read four posts about diagnosing and taken no action in a month, the diagnosis isn't what's missing. The rule of thumb: diagnose in proportion to the cost of being wrong. Cheap and reversible, just try it. Expensive, slow, or hard to undo, find out what you're aiming at first.
Our own accountability
This blog dispenses advice. Dozens of posts of it. So the argument applies to us, and it would be cheap to make it without saying how we try to earn the exception.
Three ways, imperfectly. We try to name preconditions — the sleep post says outright that sometimes sleep is the symptom rather than the cause; the safety post says some deficits are material and no framework fixes them. We try to give diagnostics rather than prescriptions, which is why so many posts end in a question you answer about your own life instead of a protocol you execute. And we try to state what the evidence doesn't show, because a citation used past its actual claim is the same failure mode as a testimonial — a real thing stretched to cover a case it never addressed.
We don't always get it right, and the record shows it — several posts in this series have been corrected after the fact when a claim turned out to be stretched past what its source actually said. The honest position is that this frame is a vocabulary for noticing, not a validated instrument, and when it stops helping you notice things you should put it down. That sentence is in a lot of our posts on purpose.
The one piece of advice we'd defend as close to general is the one this post is about, and even that is really a procedure rather than a prescription: find out what's actually thin before you spend six months feeding something else. The expensive failure isn't bad interventions. It's months wasted on good interventions aimed at the wrong target.
Where NexTier fits, briefly
NexTier is designed around this ordering — the assessment is built to come before goal-setting, because a goal chosen without a reading is a prescription without a diagnosis. That's the whole product thesis in one sentence, and it's the reason the assessment reads all eight tiers instead of asking what you'd like to work on.
If you want the structured version, the assessment is the front door. No product required — the whole thing is one instruction: work out which need is thin, then go looking for advice aimed at it. And when the next person tells you what changed their life, believe them — and then ask what would have to be true about you for it to change yours.
—
This post is part of a series on personal development organized around what we call Maslow's extended hierarchy — our synthesis of his later work. Sources: T. D. Wilson & E. W. Dunn, "Self-Knowledge: Its Limits, Value, and Potential for Improvement" (Annual Review of Psychology, 55, 2004); T. Theeboom, B. Beersma, A. E. M. van Vianen, "Does Coaching Work? A Meta-Analysis on the Effects of Coaching on Individual Level Outcomes in an Organizational Context" (Journal of Positive Psychology, 9(1), 2014). This post is educational content, not medical advice.