In Part 1 I argued that the numbers on your report divide into gauges you can only read and levers someone can actually pull, and that most dashboards are built almost entirely from the first kind.
Suppose you fix that. You run the three questions, find the buried metric, and rebuild the page so the lever gets the top-left tile.
Then you present it. Everyone nods. It goes beautifully.
And nothing happens.
The failure is never in the analysis. The analysis was fine. The problem is that a lever is not a mechanism. It is a person, and that person has a calendar, a bonus structure, a reputation, and (in their mind) a fairly good reason to leave things exactly as they are.
The Meeting Doesn’t Make the Decision #
Start with the uncomfortable structural fact. By the time you are standing in front of the room, most of the decision has already been made — in corridors, in one-to-ones, in the week before. The meeting ratifies. It rarely decides.
So if you only show up for the meeting, your beautifully rebuilt report is arriving after the vote.
The people in that room are not one audience. They are four, and you have almost certainly been optimising for the wrong one.
The requester asked for the report. They talk to you, they answer your emails, they are the reason you have requirements at all. They usually cannot say yes.
The decision maker can say yes and make it stick. Sometimes this is the requester. Often it is not, and the gap between those two is where a great many well-built reports go to die.
The champion argues for it when you are not in the room. If you cannot name yours, you don’t have one, and that is worth knowing before you present rather than after.
And then there is the silent veto. This is the one that matters.
The silent veto cannot approve anything. They can, however, quietly not do it. They agree in the meeting. They ask no questions — that is the tell — and then adoption simply never happens. No confrontation, no objection you could have answered. The thing just doesn’t take.
You find them by asking one question: who has to work differently on Monday?
That person is usually not in the room.
Every Yes Has a Price #
Here is the part that data people systematically fail to price.
If the data says X and someone’s bonus says Y, Y wins. Every time. Not because they are irrational or dishonest, but because they are measured on Y all year and your report is one input on one Tuesday. You will not fix that with a better chart.
Take the freight company from Part 1. The lever was failed pickup attempts per depot, and the recommendation was to reschedule route windows on the worst performers. Sensible. Evidence-based. Now price it:
- The operations director who asked for the analysis: costs nothing. She gets a win.
- The depot managers: rebuild driver schedules, absorb the complaints, explain the change to customers who liked the old window.
- The planner who designed the current route windows two years ago: has to watch his work get unpicked in a meeting.
That last one never appears on anybody’s stakeholder slide, and it is frequently the one that kills the project.
The currencies are worth naming, because people reach for the first two and stop: time, budget or headcount, control, status, face, and risk. Face and risk are the ones technical people never count. Who has to admit they were wrong? Who carries the blame if this doesn’t work?
Nobody in that list is behaving badly. They have simply been handed a bill that nobody told them was coming.
Why “Here’s What It Can Do” Fails #
The default opening for a report walkthrough is a feature tour. This new view lets you drill from depot to route to driver, with failed attempts alongside on-time percentage.
What the audience hears is: here is more work.
There is a well-worn explanation for why the reframe helps. Prospect theory holds that losses loom larger than equivalent gains [1], and status quo bias describes the pull of the current arrangement even when a better one exists [2]. Lead with the loss rather than the capability, and the same report becomes relief rather than request.
Loss aversion has taken some fire, and I should say so. Gal and Rucker argued in 2018 that much of the evidence is better explained by plain inertia [3]. A real critique, worth knowing — and it doesn’t change the advice, because both readings point the same direction. Their current process is known and survivable. Yours is theoretical. You are arguing against a default, not presenting to a blank slate.
So the opening becomes: we’ve rescheduled nothing on the twelve worst routes for two years, and it cost us four thousand failed pickups last year. If we decide in October we lose another peak season.
The first version is a feature list. The second is a bill.
With one hard constraint attached: the number has to be real. Loss framing without an honest number is fear-mongering, and audiences can smell it from a considerable distance. No hypothetical disasters, no “could cost us millions.” If the honest number is small, say the small number.
Objections Are Not All the Same Objection #
When people push back, it sounds similar in the room, and it isn’t. There are roughly four kinds, and technical people answer all four with more methodology.
“I don’t trust this number.” The only one methodology actually addresses. Show your working, name the source.
“I can’t act on this.” A capability problem. Make the action smaller.
“This makes me look bad.” Status. Nobody ever says this out loud, which is why it is the hardest. If someone starts asking increasingly detailed questions about your data model, check whether the real objection is this one — and give them the win, or make the ask privately.
“We tried this in 2019.” History. Name the old attempt before they do.
Halve the Ask #
Which brings us to the last move, and the one I have found changes outcomes most reliably.
Most report presentations end with “any questions?” That is not a close, it is an abdication — I have written about that before in the context of conference talks, and it is worse in a business meeting, because the report was supposed to produce a decision.
Write the ask as a sentence with three parts: [name] does [what] by [when]. Then halve it.
Restructure the route network → change the windows on the twelve worst routes → trial new windows on three routes for six weeks.
Each step down is easier to say yes to and still moves something real. The top one needs a committee and a quarter. The bottom one needs one depot manager and a phone call. And the small yes has a property the large one doesn’t: once someone has agreed to the trial, they have a stake in it going well. The foot-in-the-door literature has been describing that mechanism since 1966 [4].
Then cut every hedge out of the sentence. Kind of, maybe, I think, we should probably consider. If your analysis supports the ask, the hedge is not modesty. It is handing them a reason to say no.
The Thing I Keep Relearning #
Every time Valerie and I rebuild this workshop, I end up moving weight in the same direction. Away from the artefact. Toward the people.
Which is uncomfortable, because the artefact is the part I started with, the part that I was sure I was the most comfortable with.
But a report is not an information delivery mechanism. It is an argument aimed at a named person who has to work differently on Monday, and who has an entirely rational set of reasons not to. Find the lever and you have done the technical half of the job.
Then go and find whose hand is on it.
Join the Conversation #
Who was your silent veto? I am particularly interested in the ones you only identified afterwards — the person who never objected once and quietly ensured nothing changed. Find me on LinkedIn or BlueSky.
References #
[1] Prospect Theory: An Analysis of Decision under Risk - Daniel Kahneman & Amos Tversky, Econometrica (1979)
[2] Status Quo Bias in Decision Making - William Samuelson & Richard Zeckhauser, Journal of Risk and Uncertainty (1988)
[3] The Loss of Loss Aversion: Will It Loom Larger Than Its Gain? - David Gal & Derek Rucker, Journal of Consumer Psychology (2018)
[4] Compliance without Pressure: The Foot-in-the-Door Technique - Jonathan Freedman & Scott Fraser, Journal of Personality and Social Psychology (1966)
Photo by mali maeder: https://www.pexels.com/photo/close-up-photography-of-round-brass-colored-ship-throttle-device-53797/