The 20-minute version
It's Monday, and the team is trying to work out why sign-ups fell last month. You noticed something on Thursday. The form doesn't load properly on a phone. You start putting the sentence together, and then other thoughts arrive. Someone must have checked that already. If it mattered, one of the senior people would have raised it. You might be wrong about what you saw.
You say nothing. Ten minutes later a colleague says, "Could it be the form on mobile?" The room turns to her, someone checks, and she's right.
You knew the answer on Thursday.
There are other versions. A job advert open in a tab for four days, and you meet most of the list, but the two lines you don't meet are the ones you keep rereading, and on Friday you close the tab. An email written in fifteen minutes and then fifty more spent changing "I think" to "I believe" and back, before you forward it to a colleague with "does this sound OK?" Two years of drawings in a folder under the bed, and "oh, they're not really anything" when a friend asks to see them.
Each went the same way. A doubt arrived, as a thought like "who am I to say this?" or just a drop in the stomach. Within a second or two you'd done something about it. You stayed quiet, closed the tab, rewrote, checked, kept the folder shut. I'll call that stepping back, because you were on your way towards something and you withdrew.
The doubt and the step back feel like one event. They're two. The doubt arrives without your permission. The step back is something you do, and I think it's a habit.
The usual explanation is "I lack confidence", as though you were born short of it. But your doubt is selective. You probably calm furious customers, drive in strange cities or nurse a sick child through the night without a second thought. Doubt that turns up only in particular places, around particular people, is more likely learned than built in.
Habits last because they pay. When you step back, you skip a moment in which you could be wrong in front of people. Nobody can fault a point you didn't make or reject an application you didn't send. The extra checking sometimes catches a mistake. And the doubt feels like modesty. Each payment is smaller than it looks, and each has a cost you've stopped counting, like the colleague who got the credit and the job somebody else now has.
I'm not asking you to be certain, or loud, or to stride into rooms. Plenty of loud, certain people are wrong. And doubt isn't always mistaken, because sometimes there's a fault in the work. The aim is accuracy, a view of what you can do that matches the evidence, which I expect is better than your view. You've been told to believe in yourself, and it hasn't worked, because belief can't be produced on request. So nothing here asks you to feel confident. If a chance comes up tomorrow to say the thing or put your name down, though, take it.
Doubt isn't always homemade. Some people work among others who do question them more, because of their age, sex, accent, colour or background. If that's you, part of your doubt is an accurate reading of the room and not a fault in you.
And ordinary self-doubt attaches to tasks. If what you have is a sense of being worthless in general, with low mood, poor sleep or a loss of interest lasting more than a couple of weeks, that may be depression, and it's worth seeing a doctor. If you're having thoughts of harming yourself, contact a doctor or a crisis line today, or the emergency services if you might act on them.
The near miss
Slow the Monday meeting right down.
Just before you decided not to speak, your chest was tight, or your face warm, or your heart quicker than a person sitting in a chair needs. You pictured the faces turning towards you. Then you let it go, and within a breath or two the tightness eased. You sat back and the meeting carried on without you. That easing is the payment. It comes in two seconds, every time, while the annoyance with yourself comes an hour later, and habits are built by whatever comes first.
Look at what you were relieved of. Nothing bad had happened. You were relieved of a prediction, something like "I'll say it, it'll be wrong or obvious, and they'll think less of me". Stepping back didn't test it. It removed the only situation in which it could be tested. Your mind then files a near miss. You nearly spoke, you didn't, nothing bad happened, and the credit goes to keeping quiet. Next Monday the prediction turns up with one more near miss behind it.
Therapists who treat anxiety call moves like this safety behaviours, and the standard account is that they keep fear alive because the person never finds out the outcome wouldn't have happened. The account is widely used, though the experimental evidence is thinner than you'd expect, and applying it to ordinary self-doubt is my step. I think it fits. If it does, then a good share of next month's doubt is being manufactured by this month's step backs.
There are plainer costs.
- No evidence. Your opinion of yourself rests on a record with all the hard attempts missing.
- No practice. Fifteen years of stepping back leaves you with the practice of someone who's spoken up a dozen times, so it feels harder, and you take that as proof.
- Other people fill the blank. Most won't decide you're thoughtful. They'll decide you had nothing to add, and stop asking.
- Hours. The fifty minutes on the email were real minutes.
- Shrinking wants. After enough years you stop seeing the adverts, and from the inside that feels like being realistic.
Sometimes you'd have been wrong. Say two of every ten points you hold back would have been mistaken or old news. The habit saves you those two small embarrassments and costs you the other eight. And think of the last time a colleague raised something that turned out to be old news. You gave it about thirty seconds' thought and have never held it against them.
Holding back can be a judgement. A judgement weighs the situation and can come out either way. If yours has come out the same way every time for years, the feeling is deciding and you're supplying the reasons.
Try this. Take the last three times you stepped back and write, in one sentence each, what you predicted would happen if you'd gone ahead. Many people find they can't. The prediction is just "it would have been bad". A prediction that never says what it expects can't be shown wrong, and that's a large part of how it has lasted.
The belief holding the rest up is "if I doubt it, there must be a reason". Feelings often have noticed something. Unease on a dark street is worth listening to. Sometimes doubt about your work is too, because you did the figures in a rush on Wednesday.
But look at when your doubt is loudest. When the task is new, the stakes are high, someone senior is watching, and the thing matters to you. When you're tired, or it's a few days after a mistake. None of that measures ability. It measures exposure and caring, which is why you often doubt most the work you prepared most carefully. A Dutch team in the 1990s found that anxious patients rated a written scenario as more dangerous when the character in it felt anxious, even when nothing in the facts was dangerous. Their title was "If I feel anxious, there must be danger". I think self-doubt runs on the same inference, "I feel inadequate, so I must be", though the study was about anxiety and the comparison is mine.
So go by the content of the doubt, which comes in two kinds.
A specific doubt names something you could check. "I'm not sure the March figure includes refunds." "I haven't tried this page on a phone." It points at the work and can usually be settled quickly.
A general doubt names you. "I'm not good enough for this." "Who am I to say?" Nothing can be looked up, so no amount of checking settles it.
When a doubt arrives, ask it "what exactly?" If it answers with something you could test, treat it like a note from a careful colleague and go and check. If it answers with an adjective about you, it has told you nothing about the work. Watch for one trap. General doubt dresses up as a stream of small worries, the figures, then the formatting, then the figures again. A specific doubt, once checked, stays checked. If the check brings no relief, the work isn't the problem.
And everyone else seems so sure. They say "clearly" and "obviously" and nobody's heart seems to be going. But you're comparing your insides with their outsides. In one set of experiments, people giving short speeches thought their nerves showed much more than their audiences said they did, and speakers who were told this beforehand were rated as giving better speeches. Impostor feelings are common, too. A 2020 review of 62 studies found them in men and women, students and people well into their careers, though the measuring was rough enough that I wouldn't lean on any figure. Meanwhile studies of student teams found that people who overrated their ability were seen as more competent and given more influence without being any better at the task. The colleague who says "obviously" may know no more than you.
From the outside, a person with no doubt and a person who has doubt and speaks anyway look exactly the same. You can't arrange to have someone else's insides. You can do what she did.
A ticket you never check
The other defences are quicker to take apart. Each one is partly true.
"Doubt keeps me sharp." By Tuesday night you know the slides. On Wednesday you go through them three more times and get to bed at one. Thursday goes well, and you think, good thing I put the work in.
Preparation matters. Julie Norem and Nancy Cantor studied students they called defensive pessimists, who thought through what might go wrong and performed as well as optimists, and did worse when reassured. If that's how you get ready, keep it.
But its useful part converts dread into a list of particular problems, and a list has an end. The Wednesday run-throughs weren't answering anything. They were making a feeling go away, and the sign is that you couldn't say what the next hour was meant to fix. Then the success gets credited to the over-preparation, so you gain nothing from it for next time.
A standard is about the work. A doubt is about you. A pilot runs the same checklist before every flight whether she feels sure or not, and when it's finished she flies the plane. So write down what done looks like before you start, two run-throughs and three figures checked, and stop there.
"If I don't try, I can't fail."
Closing the tab didn't protect you from failure. You still don't have the job. It protected a possibility. "I could have got it if I'd gone for it." It's a lottery ticket kept in a drawer so it can go on being a possible winner. An unchecked ticket pays out exactly what a losing one pays, and you have to keep not looking at it. Apply and get turned down, and you lose an evening and have a bad day or two, and you learn something. Don't apply, and you lose the same job and learn nothing, and the open question gets heavier over the years.
The disappointment still comes, slowly. Thomas Gilovich and Victoria Medvec found that over a lifetime people more often regret what they didn't do, and a large repeat in 2022 mostly found it again, with weaker effects and caveats. Expecting the worst doesn't cushion much either. In studies by Sarit Golub, Daniel Gilbert and Timothy Wilson, people who expected a bad result felt worse while waiting, and once it arrived, what they'd expected made no measurable difference. Most rejections come from things you never see, an internal candidate or forty other good applicants. So aim to be someone who finds out. Watch for half-tries too, the application sent at the last minute, the idea introduced as "only a rough thought".
"At least I'm not arrogant." You know a man who's never in doubt, talks over people and explains why his mistakes weren't his. Your doubt feels like the wall between you and him.
Humility means being accurate about your limits and open to correction. Accuracy works both ways. A humble person can say "I don't know much about tax law" and "I'm good at explaining technical things" in the same flat tone. Bathroom scales that read five kilos light every morning aren't being modest. They're wrong.
Doubt does worse at openness than you'd think, because criticism lands as confirmation of the general doubt, so you dread feedback and put off asking. His real problem is that he's stopped listening. You can rate yourself accurately without stopping.
Self-doubt also keeps attention on you. In the meeting the subject was sign-ups, and for a minute it became how you'd look. When careful people stay quiet, decisions get made by whoever's left. Yes, some workplaces penalise people, and women especially, for talking up their strengths. But between silence and showing off there's a great deal of room, and pointing out a broken form is nowhere near boasting.
"They were only being kind."
Your manager says the report was excellent, exactly what the board needed. Before she's finished, your mind is on it. She probably skimmed it. The bar is low. Anyone would have done the same. A month later she says one section of another report was a bit unclear, and that goes straight in. You don't ask whether she skimmed it, or whether one unclear section out of twelve is good. You think about it on the way home and again on Sunday.
Everyone feels bad things more than good ones. A 2001 review was titled "Bad is stronger than good". But you're running two sets of rules. Praise is cross-examined and criticism admitted without a question. Success has a dozen causes, luck, help, an easy task, hard work, being liked, and failure has one, which is you. A shop that wrote down every refund and never rang up a sale would show a loss every year. Your record has been kept that way.
Eleven years of luck under different managers is a strained explanation. Hard work is an ability. And if anyone could have done it, it's fair to wonder why they asked you.
Weigh praise like any evidence, by how specific it is and whether they knew the work, and weigh what people do most. Did the client renew? Did you get the next hard project? Start a record, one plain line per entry, criticisms included. "Client renewed for a second year." "Section 3 unclear, rewrote it." And when you're praised, say "thank you" and stop.
"Can I just check this with you?" A second pair of eyes catches things, and asking people who know more is a fine habit. The line falls between information and reassurance.
Before you ask, say what you think the answer is. If you know she'll say "looks fine", and you'd send it as it stands if nobody were around, you're after reassurance. It works for a minute, the credit goes to her, and your own judgement never gets tested alone. Clinicians describe the same loop in anxious and obsessive problems, and carrying it across is my step. At work, people whose work always comes with "can you just check?" get seen as people whose work needs checking. At home, years of "you choose" leave you not knowing what you'd like. So ask specific questions when review is needed. For a fortnight, write down what you expect before you ask anyone to check, and compare. And when someone asks where you'd like to eat, name a place, even if you add "but I'm easy".
"I'll do it when I feel ready."
Think of driving. In your first lesson you gripped the wheel while a stranger hovered over a second brake. The ease came somewhere in the first few hundred hours. At no point did you feel like a driver first and then start driving.
Albert Bandura found that belief in your ability to do a particular thing comes most strongly from having done it, and that being told you can is weak and fades. That's why "believe in yourself" lasts an afternoon. Waiting to feel ready means waiting for the result of something you aren't doing.
So measure readiness by something you can check. List what the role needs, mark what you've already done in some form, and ask of the rest whether you could learn it in the first few months. Nobody arrives able to do all of a job. This isn't faking. "The form seems to be broken on mobile," said with your heart thumping, is a perfectly honest sentence.
The first ten minutes
Draw a map of your doubt. Where is it strong? At work and not at home, with figures and not with people, around older men or a certain kind of education? Where is it missing? The map usually has a logic, and the logic points somewhere.
Perhaps the ninety per cent on the test got "what happened to the other ten?", meant kindly. Perhaps you were compared with the clever sibling, or you were the clever one and a struggle threatened the label. Perhaps a teacher read your work aloud and the class laughed. Perhaps you were the first in your family at university and concluded that the lack was in you, when it was only in what you'd been shown. In each case stepping back was sensible at the time. The trouble is that it never got reviewed, and each step back prevents the very experience that would show it's out of date.
If the room really is the source, keep two sentences apart. "They doubt me" can be completely true while "I'm not good enough" is completely false. The record you're keeping gives you evidence that doesn't depend on the room's opinion. The remedies for the first sentence are allies, others in the same position, raising it where that's safe, and sometimes leaving. That's changing the situation, and no book method fixes an unfair workplace. One harsher case. If a partner regularly tells you your memory is wrong, your judgement can't be trusted and you'd be lost without them, the doubt is being produced on purpose. That's a form of abuse. A domestic abuse helpline is the right first call, and a therapist can help with what it leaves behind.
Here's the rule.
A doubt that names something I can check gets checked, once. A doubt that only names me doesn't get to decide what I do.
The doubt can still come along and sit in the meeting saying its usual sentence. It just loses its vote.
- Learn your step backs. Write down your two or three usual ones, the silence, the fifth reread, the "quick check", the "you choose". For a week, just notice them, even an hour late.
- Ask "what exactly?" If it names something, check it. If it names you, carry on with what you were about to do.
- Check once. If the same doubt comes back after it's been answered, it's turned into the general kind, and more checking won't help.
- Go early. Say something in the first ten minutes of a meeting, even a question, because silence sets harder with every minute. Decide which draft you'll send, the second, and send it. Put the application in the day you decide.
- Drop the apology in front. "This is probably a stupid question, but" tells people how to take what follows, and they mostly will. You needn't sound surer than you are. Just stop sounding less sure. If you're truly uncertain, say what about. "I've only tested it on my own phone" is useful. "This is probably nothing" isn't.
- Write the prediction, then the result. This matters most. Before, one line specific enough to be wrong. "They'll say it's already been checked and I'll look foolish." After, one line. "Dev checked, it was broken, fixed by lunch." Therapists call these behavioural experiments. My expectation is that most predictions fail, and the ones that come true are milder than forecast.
- Bank the good results in the same record, and read it through before anything that frightens you.
Start small, as method. A question in a small meeting, an email sent on the second draft, a preference stated at home. But start this week. Then write two or three plans for the moments you know are coming. Plans in the form "when this happens, I'll do that" have been tested in many studies, and people who make them follow through noticeably more often than people with only a good intention.
- When the team meeting starts, I'll ask or say one thing in the first ten minutes.
- When I've finished the second draft of an email, I'll send it.
- When someone asks where I'd like to eat, I'll name a place.
The first dozen times your body will behave as it always has. The feeling follows later, as the record fills. Many people who speak in public for a living are nervous every time. They've stopped treating it as a reason not to speak.
Six out of ten
You make the point, and it was dealt with last week while you were on leave. The doubt is on it at once. There, you see.
Put the event next to the prediction. You wrote "I'll look foolish and they'll remember it". What happened was "yes, we picked that up last week, thanks", and the meeting moved on in four seconds. The doubt got the small thing right and the big thing wrong, and the big thing was what had been stopping you. When a report goes out with a mistake, find the specific cause. The refund figures come from a different system and you didn't reconcile them. That's a new line on your checklist, and a fact about refund figures. A rejection often has nothing to find, so count the attempt as the success, because you found out. Even a real gap, a muddled talk, is a specific doubt on a bigger scale, and it comes with a to-do list. And if you speak ten times where you used to speak once, you'll be wrong in public more often and right far more often, and people judge the whole run.
Or you sit through Tuesday's meeting and say nothing, after deciding not to. "You can't even follow a self-help book." That's the same voice making the same move. A lapse in a habit you've had for twenty years is what changing a habit looks like. Juliana Breines and Serena Chen found that people prompted to treat a failure with understanding were more motivated to improve, and studied longer after doing badly, than people who weren't.
So ask whether there's still time. Often there is. "One thought I didn't get to raise. The form seems broken on mobile." Ask what was different. The director in the room, a tired week, minute forty and the silence set hard. Plan for that. "When the director is in the meeting, I'll speak in the first five minutes." If you keep failing at a step, drop a rung for a fortnight. Don't count streaks. If you stepped back nine times out of ten last year and it's six out of ten by summer, it's changing.
Get more help than a book, starting with a doctor or a qualified therapist, if any of these apply.
- Low mood, loss of interest or a sense of worthlessness lasting more than two weeks.
- Panic attacks, or fear of being judged so strong you avoid phone calls, eating with others or going to work.
- Checking you can't stop, taking hours of your day.
- Doubt that goes back to abuse, bullying or another experience that still intrudes on your life.
- Months of work on this with nothing moving.
For severe anxiety about being judged, which doctors call social anxiety disorder, the health guidance body for England recommends individual cognitive behavioural therapy first. If you take prescribed medication for anxiety or depression, don't stop or change it because of anything here. Talk to whoever prescribed it. If you ever have thoughts of harming yourself, contact a doctor or a crisis line that day, or the emergency services if you're in immediate danger.
Before the next meeting, do three small things. Write one "when, then" plan for your most common step back and put it where you'll see it. Start the record with three plain facts from the last year, and don't let yourself explain them away. Tell one person, in a sentence. "I'm trying to speak earlier in meetings."
The doubt will arrive on time. Ask it what exactly. If it names something, check that one thing. If all it has is its opinion of you, you know where that came from. It won't feel good, and afterwards you'll find fault with how you put it. Write down what you predicted and what happened.
You knew the answer in that meeting, and a good many others over the years. Say the next one.