If every business ran the way it's described, why does a site need to get ready for a visit?

A consultancy fields a whole team, and you get the one who's free that week — usually someone bright, two years out of university, who has read about your business but never held a chair in it. The partner you met at the pitch does not do your work. You're not buying hours. You're buying a read that only exists because I ran the departments I'm assessing — and I deliver it start to finish.

The version that reaches your desk has been through every person between the floor and you, each with a reason to round it up. If every business ran the way it's described, no site would need to get ready for your visit — and every one does. What I bring back is the part sanded off on the way up: the reason the good manager is actually leaving, the junction where the money leaks before ten o'clock, the thing everyone in the building knows and nobody will say to you — because saying it is career-limiting for them and unwelcome for you. I have no career left to limit inside your business. That is precisely what you're buying.

A written finding, not a slide handover. Every line carries a named owner and a date — if it can't be written that way it's an observation, and observations fill drawers. What's already working is protected in writing, so a review never breaks the thing that wasn't broken. Timelines are set at scoping and held; most reads land within three to four weeks, and I come back at ninety days to see what stuck.

A survey produces the appearance of having asked. People don't believe the anonymity — rightly, because a team of six is identifiable by arithmetic once results are cut by department. So it returns exactly the data the business hoped for. I don't run an instrument; I speak to your people, at every level, no manager in the room. It isn't anonymous — it's unattributable. What I'm told, I keep, and many of these people have known me thirty years. That's why they tell me the truth.

Then you'll have it in writing, with the reasoning, and you can act on it or not — your call. What I won't do is soften a finding, negotiate a line out, or let anyone pre-clear what gets said before it reaches whoever commissioned the work. A finding filtered through the people it implicates isn't a finding. If you want a review that confirms a decision you've already made, I'll recommend three consultants who do that well. It isn't the job I do.

The Entry Review. Half a day, your site, no charge. I walk your building, stand at reception at eight o'clock, stand at the parts window — then tell you three things I saw that you can't, because you're in it. Verbal, on the day, to you alone. No report, no invoice, no obligation, and I don't chase you afterwards.

It's a deliberate risk on my side, a small one on yours. If three true things about your own building aren't worth paying for, we both learn that at no cost — and I'd rather learn it then than after you've written a cheque. The door is below.

It is exactly their job, and the good ones do it well. But every one of them reports to somebody, and reporting upward is never neutral — the person telling you what's wrong is also telling you how well they've managed it. That's not dishonesty; it's the arithmetic of a hierarchy. You're buying the version your team structurally cannot give you, and the ability to compare it against what they did. If the two match, you've confirmed your management line is sound. If they don't, you've found what no internal process was going to surface. And it isn't that they haven't looked — they're running reactive, because the day makes them: a target that moved, a resignation nobody saw, a week of sickness in a thin department, four cars gone off the floor before an event, damage on a service car that wasn't there at check-in. All of it has to be dealt with today, by the same people who were going to do the proactive work. I'm doing the part the day never lets them reach.

There isn't one, and I won't pretend otherwise. This is not benchmarking. If you want your network scored against a panel of two hundred sites, buy that — it's a real product and I'm not it.

What I sell is a second opinion, from one person, on a specific plan or a specific building, delivered in weeks rather than quarters. A second opinion is supposed to be one person — that's what makes it a second opinion, not a second committee. The depth is the trade: I can't tell you where you sit against the market, but I can tell you what happens in your service reception at half past eight, because I've stood there for thirty years across five marques and over forty sites.

Judge it on the arithmetic. If a read costs a fraction of the programme it's checking, and it's wrong, you've lost a small number. If it's right, you've avoided a large one.

Because I don't report people. I report mechanisms. No individual is identifiable in anything I write — not by name, not by role where the role is a team of one, not by the arithmetic of who could have said it. Where a finding can only be described in a way that identifies someone, I describe the pattern instead or leave it out. That costs me something — occasionally the sharpest example is the one I can't use — but the alternative destroys the mechanism. Identify one person once, and nobody in that building tells me anything again, nor anyone in the next. What you receive is a system failure with an owner attached to the fix, not a person with a case attached to them. You can act on every line in public without exposing anyone.

Only if you're the subject rather than the sponsor — and here, you're the sponsor. The person who commissions me usually knows roughly what I'll find. The problem is that saying it upward, with the uncomfortable parts intact, carries a cost — it reads as excuse-making, or a bid for headcount. I carry the message you can't carry yourself, and the finding is about the mechanism, not the management. Nobody sees a verdict on the person who brought me in; that's in the terms. There's a plainer version too. Across a double-digit number of sites you can't be in every building, so you trust the reports — and the dashboards are excellent. Everyone knows exactly what the data says to do. Then service is well off target with no capacity to recover the month, and the site is staring at a loss. The data told you what happened, accurately and early. What it can't tell you is why capable people who know the job can't do it — and that answer never travels up the line intact, because it costs the person writing it something to say. Used properly, this is your instrument, not your audit: the director who commissions it arrives at the board with an independent account of what the network will do, not an opinion that gets discounted as self-interested.

Every line carries a named owner, a date, and a definition of done. If a finding can't be written that way it isn't a finding — it's an observation, and observations are what fill drawers. So Monday morning you have a list of things, each of which belongs to somebody by name and is due on a date.

Two rules govern what goes on that list. First, no recommendation that requires headcount you can't sign off. If the answer to everything is "hire someone," you didn't need me. The work is making the people you already have more effective, which usually means taking something out before adding anything. Second, what's already working gets written down and protected before any problem is named — so nobody enthusiastically fixes the thing that wasn't broken.

And I come back at ninety days. Not to present again — to find out what landed, what quietly reverted, and what I got wrong because somebody managed what I was allowed to see.