Seven percent on task. The shame is the expensive part.

The shame spiral bills you for the week, not the hour

The ADHD shame spiral is the gap between what a bad stretch of work actually costs and what you end up paying for it. Forty minutes of drift costs forty minutes. What arrives next is a verdict — today is a write-off — and the verdict takes the rest of the afternoon, then makes tomorrow morning harder to open, because opening it means facing the record of yesterday. Most writing about this is aimed at students, where the stakes reset every term. For a working adult the spiral compounds: the week has deliverables in it, other people are waiting on them, and each written-off day makes the next one more expensive to start.

Here is the specific session that made me write this down. I sat down to finish a launch page. I opened it, read two lines, and then spent most of the next hour somewhere between Reddit and a tab about mechanical keyboards I do not own and will not buy. When the session ended, the honest figure was 7% on task.

Seven percent is a humiliating number to look at. It is also a real one, and the difference between those two facts is the whole subject of this note.

A small lapse gets recoded as a lost goal

A small lapse gets recoded as a lost goal, and that recoding is the mechanism — weak will is not required for any part of it. Polivy and Herman described the pattern in American Psychologist in 1985, working with dieters: someone who breaks a rule by a little frequently goes on to break it by a lot, because once a goal reads as already blown, the remainder stops feeling worth defending. Later writers named it the what-the-hell effect. Their research is about food rather than focus, and I want to be plain that carrying it across is an analogy and not a finding. The shape is what transfers: from the inside, 7% and 0% feel identical, though only one of them is a number.

The second mechanism is quieter and does more damage, because it arrives long before the session does.

In a 2022 study in PLOS ONE, Danielle Beaton, Fuschia Sirois and Elizabeth Milne collected written accounts from 162 adults with ADHD, aged 18 to 62, describing criticism they had received from family, friends and colleagues. The authors found the criticism had not stayed outside: it "impacted their sense-of-self to a point where the criticism was internalised", leaving participants critical toward themselves. One described being unable to separate herself from negative criticism that then fuelled her own low opinion of herself.

That is what makes the adult version harder than the version aimed at teenagers. By the time you are thirty-three and reading a 7% on your own screen, nobody has to deliver the verdict. Thirty years of other people's comments have already installed a serviceable version of the critic, and it works weekends. The lapse supplies the evidence and the internalised voice supplies the sentencing, and between them the afternoon is gone.

Shame is a punishment term bolted onto variance reporting

Shame, in delivery terms, is a punishment term bolted onto variance reporting. Variance is the gap between what you planned and what actually happened, and every project ever run has some. The discipline around it is famously dull: measure the actual, compare it to the baseline, explain the gap, re-baseline, carry on. I spent years doing exactly that for banks — reconciliation work at Aon, where the entire job was finding where the number that arrived differed from the number expected, and saying so early enough to matter. Nobody has ever improved a reconciliation by feeling bad about the break.

What does break a project is making the variance expensive to report. I have sat in the 2 PM steering call where a team reported green for three weeks and then slipped a month in a single meeting. They were not lying in week one. Amber had been made costly to say out loud, so amber stopped being said, and by the time the number could no longer be hidden, the buffer that would have absorbed it had already been spent. Punishing variance does not reduce variance. It reduces reporting, and reporting is the only instrument you have.

Run that same failure on a single person and it looks like this. The day goes badly. The honest figure would be 7%. Writing 7% hurts, so nothing gets written at all, and tomorrow opens with no data and a feeling. The feeling says the whole week is going the same way, and because there is no record to argue with, the feeling wins.

Five moves that make a bad number cheap to write down

Making a bad number cheap to record is the move that does the most work here, and none of the five below need willpower on the day you need them. Willpower is the resource a bad session has already spent. Each of these is either a change to what you write, or a change to what counts as the smallest unit of work, and both are decisions you can make in advance on a good day and simply follow later. None of them require you to feel differently about the afternoon, which matters, because feeling differently on demand is the one thing nobody can schedule.

1. 01Record the number before the verdict. Write the figure first — 40 minutes, two pages, 7% — and only then, if you must, write a sentence about the day. A figure is data you can plan against. "Wasted" is a rating, and ratings compound.
2. 02Give every session a floor of one action. If the smallest unit is "open the file", then a bad session still ends on a one rather than a zero. No zero, no lost goal to abandon, and nothing for the what-the-hell arithmetic to work with.
3. 03Forgive the lapse once, in writing, then stop. One line: I lost the afternoon, and I am not billing myself twice for it. The evidence here is real but narrow — it held only for people who had procrastinated heavily — so save it for genuinely bad stretches rather than making it a daily ritual.
4. 04Re-baseline to the next smallest thing, never to the original plan. A recovery plan that restores the whole day is a fantasy that fails by three o'clock and then confirms the verdict. Re-baselining to one paragraph tends to survive contact with the actual afternoon.
5. 05Close the day with two lines of what actually happened. Not a reflection, not a journal. Two lines, so tomorrow inherits a record instead of a mood. On the bad days this is the entire point of the practice, and on the good days it costs ten seconds.

What changed

I still have 7% afternoons. That has not moved much, and I have stopped expecting it to.

What moved is the cost of one. The number goes in the record, the baseline redraws to something small enough to survive the evening, and Thursday no longer inherits Wednesday's verdict. Nine planners went blank after page three because each one asked me to be a person who had good days. The card asks me to be a person who writes down what happened, and I can manage that on a bad day, which is the only test that matters.