The review you need is never available when you need it

This is the ordinary constraint that shapes more work than any technical limitation. Not that scrutiny is unavailable in principle, but that it is unavailable now, and now is when the document is being finished.

So it goes out unreviewed, or reviewed by whoever happened to be free rather than whoever would have been useful. The weak assumption survives to the meeting, where it gets found by someone in public rather than by you in private, which is the same discovery at roughly ten times the cost.

What has changed is that you can now build a reasonable approximation of that scrutiny and have it available at eleven at night, which is when most documents are actually finished.

The specificity problem#

Most attempts at this fail, and they fail for the same reason: the instruction is a job title.

Tell a system to act like a finance director and you get an encyclopaedia entry with opinions. It knows what finance directors care about generically and will return generically correct observations, all of which you had already thought of. It is a party trick, it wears off in a fortnight, and people conclude the idea does not work.

The difference between that and something useful is temperament rather than role. Not "a CFO" but someone who has watched businesses fail while profitable on paper, and who goes to cash before anything else. Someone visibly impatient with the phrase "once we scale". Someone who asks what happens at thirty per cent of your projection rather than at a hundred, who wants best case, worst case and most likely before looking at the headline, and who opens with "let me play devil's advocate" because that is how they soften a question they are going to ask regardless.

Give it priorities in a definite order, because a critic without ordering simply lists everything and leaves you to work out what mattered. Give it a manner, because how criticism arrives determines whether you act on it. And give it boundaries, so that it is not obstructive as a personality and you can distinguish a real objection from reflexive negativity.

That collection is a colleague. The job title was a costume.

What a small team looks like#

Three or four well-drawn perspectives do almost all of the useful work, and I would take three sharp ones over fifteen vague ones without hesitating.

The commercial mind attacks the money and the timelines, and finds the estimate that has quietly become a fact somewhere between the second and third draft.

The risk and compliance perspective notices what you have accidentally promised. Guarantee, always, never and ensure are load-bearing words that people deploy casually, and someone whose reflex is to ask whether you could actually be held to that catches things while they are still free to fix.

The brand or communication perspective reads it as a stranger would, notices where the writing has gone corporate, and finds the point buried in the final paragraph where you put it after working out what you thought.

The fourth, if you want one, is a strategic alignment check configured with your organisation's actual stated priorities, evaluating whether your case connects to them. Unglamorous, and it quietly saves more proposals than the other three combined, because proposals mostly fail on relevance rather than quality.

The part I did not expect#

There is a benefit that appears after a couple of months and is worth more than any individual review.

Keep the feedback and patterns emerge. If the commercial perspective questions your timelines every single time, that is not a characteristic of the tool. That is a systematic tendency in your thinking, now visible across a dozen documents in a way no annual review would ever have surfaced.

I have found this uncomfortable in the way useful things usually are. It turns out I am consistently optimistic about how quickly people will take up something new, which I half knew about myself and had never seen laid out as evidence.

Where the boundary sits#

Everything here rests on one distinction, and the whole thing collapses without it.

These perspectives prepare work for your judgement. They do not exercise judgement. They find the weak assumption; you decide whether to strengthen the argument, change direction, or proceed knowing the risk because you have information the critic does not.

That last option matters more than it sounds. Sometimes the objection is wrong and you are right, and being able to articulate why is precisely the confidence the exercise was meant to produce. A critic that always wins has become an authority, and you did not want an authority, you wanted a rehearsal.

The discomfort is worth examining#

Some people find this idea unsettling, and I do not think that reaction should be brushed aside, because it usually contains something.

Part of it is a reasonable objection to outsourcing judgement, which is answered by the boundary above. Part of it is a fair worry about what happens to junior people if the easy review work disappears, and I do not have a clean answer to that one.

But part of it, if I am honest about my own reaction when I started, is that having your thinking taken apart is unpleasant regardless of what is doing the taking apart. Being told your business case has a hole in it at half past ten at night is not enjoyable. It is considerably less unpleasant than being told at ten past nine the following morning by someone whose opinion of you it will affect.

Everyone benefits from this, including the person whose scrutiny you were trying to borrow. Nobody enjoys being handed a document with obvious problems that the sender could have found alone. Fixing those before they arrive is not a substitute for the relationship. It is basic courtesy, applied at a scale that was not previously available.