Essay · April 2026

False positive alignment

Agreement on the label, not the mechanism.

I have walked out of deep, rich conversations with my teams convinced we were aligned. We had talked through the problem from multiple angles. We had broken it into explicit layers: the work we would take on, how we would prioritize across teams, who would be accountable for each piece. We had agreed on a direction. We had even agreed on the phrase: we would deliver improved insights.

Two weeks later, three people brought me three different things.

Each had done careful work. Each had executed on what they thought we had agreed. We had agreement at the level of direction and phrase. We had agreement on some of the layers I thought we had broken down. The mechanism underneath those layers had stayed unresolved, and each person had filled it in differently as they walked out of the room.

This is what I call false positive alignment. Agreement is expressed on direction, without operational definition of the mechanism underneath. The meeting ends before those gaps can surface. What remains is rhetorical alignment: real agreement on the words, no shared definition of what the words mean in practice.

The gap has been measured. In a 1990 Stanford experiment by Elizabeth Newton, "tappers" were asked to tap the rhythms of well-known songs on a table while "listeners" tried to identify them. The tappers, who could hear the melody in their own heads, estimated that listeners would correctly identify about fifty percent of the songs. The actual rate was two and a half percent. The tappers had so thoroughly absorbed the melody that they could not feel what it was like to hear only the tapping.

Senior leaders do the same thing with their own strategic decisions every day. They come out of a meeting hearing the full melody in their heads: the trade-offs they considered, the alternatives they rejected, the constraints they were optimizing against, the things that would make them reverse the decision. What reaches the room is the tapping. Phrases like "improved insights," "customer-focused," "efficient" become the words the team agrees on, while each person silently fills in a different mechanism behind them.

Systems-thinking research by Derek and Laura Cabrera at Cornell names this at a deeper level. Meaning is produced through four cognitive acts operating together: the distinctions people draw between things, the systems of part-whole relationships they see, the relationships they perceive between parts, and the perspectives from which they observe. Two people with the same information can walk away with different meaning if any of those four are not aligned. Rhetorical alignment does none of that work; operational definition of the mechanism does.

A Harvard Business Review study of more than a thousand companies found that information-flow quality and clarity of decision rights predicted successful strategy execution more strongly than any other factor: more than structure, incentives, or the quality of the strategy itself. The variable that moved execution the most was whether people understood what decision they were actually supposed to make, and on what basis.

Building shared context takes longer than distributing information, and leaders resist the time cost. The minutes spent in the room defining the mechanism operationally save the weeks you would otherwise spend downstream, discovering the mechanism was never shared.

The correction is a discipline I now run at the end of every conversation that produces a material decision. Each person in the room says, in their own words, what they believe was decided and what they will do first. If the mechanisms they describe match, we have alignment. If they diverge, we have rhetorical alignment only.

The three people who brought me three different things had done careful work. They had gotten to the answer they thought we had asked for. None of them had made a mistake. The room had produced rhetorical alignment, and the cost had landed downstream. It took three more conversations to make the mechanism operational.

Most instances of false positive alignment never get caught. The work ships, the misalignment gets absorbed into next quarter's priorities, and no one goes back to ask why the three versions diverged. The cost is paid; the mechanism never gets named.

Name the mechanism. Have each person repeat it back. Anything less is rhetorical alignment.