The Compass Does Not Show the Way── Before Correctness Becomes One-Way

01He Only Said "I'll Fix It"

I was reading a younger colleague's review comments. They concerned a promotional piece for a new product. Take the way the numbers were shown. The headline carrying the main result said "reduced the event rate by 30%." The number is correct. Five percent became 3.5 percent — that is the relative reduction (= the percentage drop measured against the original figure). But the absolute difference of 1.5 points (= how many points the actual event rate fell) sat in small type in a footnote. My colleague went straight at it. "Present relative and absolute side by side, with equal weight. Here is the relevant section of the creation guidelines; here are past cases where this was flagged." The comments on the choice of cited literature and the placement of safety information had the same tone. I'm using only the numbers as an example here, but every point was correct. More careful, if anything, than what I would write.

And yet, in the meeting with the person who had made the material, that correctness never reached him. He pushed back once. "This part is essential to the product strategy — we can't lose it." My colleague answered by stacking up more grounds. The answer was right. The man said nothing further, only, briefly, "I'll fix it." When he left the meeting room, there was no trace of acceptance on his face.

For him, fixing it in silence may have been the surest road to protecting the launch date. Keep pushing back and the review drags on. You start the next material at a disadvantage — even if that isn't actually true, it would be no surprise if he felt it was. Silence is resignation, and at the same time a reasonable choice. Which is exactly why the reviewing side misreads silence as settlement.

I recognize this scene. Years ago, I was doing the same thing. The comment I wrote about in an earlier installment — the one where I announced a conclusion without showing my grounds — was precisely this. The point gets through. The material gets fixed. Nothing remains inside the other person.

02Rightness Never Runs Out, But It Doesn't Reach

Rightness is a strange thing: use it as much as you like and it never runs out. The standards don't change; the data doesn't change. So a reviewer can deploy the same rightness again and again.

But whether it reaches anyone is a different matter.

A line like this often fills a review comment field: "Not permitted, per Section X of the creation guidelines." That's all. Sufficient for the person who wrote it; to the person receiving it, no different from a notice pasted on a wall. If you were to add one line, it would be this: "Tell me why you chose this graph. If I understand the difference you want to convey, we can look for another way to show it together." The former is a verdict; the latter is a question. For a long time, I wrote nothing but verdicts.

Psychology has a term, "perspective-taking" (= the effort to see things from the other person's position). The twenty-five experiments published in 2018 by the psychologists Eyal, Epley, and colleagues shook our faith in it. In experiments that asked people to guess another person's preferences and feelings, merely imagining the other's perspective barely improved accuracy. What rose was confidence, nothing more. Accuracy improved not when people imagined, but when they actually asked.

Transpose this into the review room and it stings. We imagine the circumstances of the people who make materials. How the sales rep intends to use this figure at a briefing, what they want to say when it's placed next to a competitor's material. Behind that imagining there must also be the maker's own accumulated experience. That 30% headline, too, may have come from experience: "The relative reduction is what carries the clinical meaning intuitively; it's used routinely in conference presentations." The maker has a needle too (= a personal sense of judgment that points the way forward). A needle pointing in a different direction from the reviewer's, honed by knowing the product and the customer through and through.

And still we pile correct comments on top of imagination. We don't ask. Why this structure? What did you want to convey when you chose this graph? Rightness thrown without asking, however right it is, travels one way. The revised version circulates, gets approved, and no one ever speaks of that material again. Only the revision history piles up, and nothing grows.

03What Needs Breaking Is Not the Needle but How the Compass Is Used

Every reviewer carries a compass (= an instrument that tells you which way to go) inside. A needle for what is right and what is dangerous, polished through experience and study. I want to treasure that needle. It is the makers who first create and raise the materials, but at the final gate that protects a material from danger, a reviewer without a needle cannot stand.

Yet sometimes I think: this compass has to be broken once and rebuilt. Starting with my own.

Breaking it doesn't mean changing the direction the needle points. What needs breaking is the way the compass is used. The assumption that if you show the direction, the other person will start walking. The habit of treating the compass as something you peer into alone. To borrow the words of the philosopher Martin Buber, an exchange confined to the comment field stays at "I–It" (= a relation that treats the other as an object to be processed). There is no "I–Thou" (= a relation that faces the other as a person) in it. Once, after a rejection, another material maker came to ask me, "Was that a rule violation, or your personal opinion?" What that person was looking for, I think, was the latter relation.

Reading the compass together is nothing grand. Before pressing the reject button, make a ten-minute phone call. Afterward, record what was discussed in the system's comment field, so it doesn't remain a verbal promise. If there are ten comments, talk through just the first one in a meeting before writing all ten in. If the first one reveals the maker's design thinking, the wording changes for half of the remaining nine. Sometimes a comment disappears altogether. And the fixes come faster after a phone call, not slower. In my experience, the materials I rejected in silence are the ones whose second and third rounds drag on longest.

A compass tells you the direction. It doesn't tell you the road. The road is something you walk with the other person, map spread open. Remaking a tool for one into a tool two people read together — that, I think, is what breaking it and rebuilding it means.

04Start from Where the Two Needles Diverge

So what would a compass of rightness that two people can read together look like? Honestly, I don't have the answer yet. But I've picked up some clues.

The first: ask before writing. And the person to ask is the one who made the material. Before turning to the comment field, walk over to their desk or pick up the phone and ask, "Could you tell me why you chose this graph?" Eyal and colleagues' experiments have a sequel of sorts. Imagining doesn't raise accuracy; it only inflates confidence. Which means the comments written without asking are the ones written with the most conviction. Last week, again, I wrote without asking. The very strength of certainty in a finished comment may be the proof that I didn't ask. When the needle swings, ask the maker once before writing. The only entrance more reliable than imagination seems to be a question.

The second: ask again after the fix. I wrote in installment 26 about what happens when someone fixes things in silence. A quiet review is not proof of success. Did they fix it because they were persuaded, or because they had no time? Reading faces is, in the end, just imagination continued.

And the third. Disclosing where I stand — "up to here is the rule, from here on is my own view," which I wrote about in installment 24 — was only an entrance. Even after disclosing, my needle and yours diverge. Take the warning text for side effects. As the reviewer, I want it bold, boxed, the same size as the body text. I'm afraid of it being missed. The maker thinks: a page where warnings dominate won't be read through, and a material no one reads, however correct, means nothing. Both of us are facing the patient. Facing the same way, and diverging. Do you see that divergence as an error to be eliminated, or as the real subject to be discussed? I've come to think it's the latter. The gap between the two needles is exactly what we should truly be talking about when we talk about a material. Though whether to show the needle at all is the other person's freedom. Demand it, and we're back to one-way traffic.

My young colleague's rightness is the real thing. Which is why I hope it will someday become a rightness that can walk alongside the other person. That is a demand on my colleague only after it is a demand on me — twenty years in this work, and there are still days I peer into the compass alone. I still haven't asked "why" of the man who only said "I'll fix it." What he had in mind when he set that 30% headline. It's a matter a ten-minute phone call would settle. That's where to start.

Break the compass, and rebuild it. As many times as it takes. Until it becomes a needle two people can confirm together. Whether that rebuilding ever ends, I don't yet know.