Stop Presenting Data. Present the Decision It Supports.

Strong analysts routinely watch weaker analysis get adopted over theirs. The usual explanation offered is politics. More often the cause is that the weaker work arrived with a clear story and theirs arrived as evidence, leaving the audience to construct the conclusion themselves.

An audience will not do that work. Not because they are incapable, but because in a thirty-minute meeting the person who supplies the interpretation controls the decision.

One chart, one sentence

Every exhibit should have a single message, and that message should be the title. Not "Q3 Revenue by Segment" but "Enterprise growth is masking a decline in mid-market." The descriptive title makes the reader find the point; the assertive title makes them evaluate your claim, which is what you want.

If you cannot state an exhibit's message in one sentence, it is a reference table. Move it to the appendix.

Show what changed, not everything you have

Analysts include the full picture because they are worried about being accused of selectivity. The result is dense material where the actual finding is one of eleven visible things.

Show the comparison that carries the argument, note what you excluded and why, and keep the rest available. Volume is not credibility; it usually reads as an unwillingness to commit to a conclusion.

An audience does not remember your analysis. They remember the sentence you gave them to repeat to someone else.

Give the number a scale

A finding stated in isolation lands weakly. "Churn is 14 percent" invites no action. "Churn is 14 percent, against 9 percent last year and roughly 7 percent for the two competitors we can observe, which is about eleven million in annual revenue" produces a decision.

Three anchors do most of the work: what it was before, what it is elsewhere, and what it is worth.

Name the uncertainty precisely

Vague caveats — data quality issues, directional only — teach an audience to discount everything you present. Precise ones do the opposite: this holds for enterprise accounts, where we have clean data; the mid-market figure is based on a partial sample and I would not act on it yet.

Specific limits make the rest of the analysis more believable, not less.

End with what you would do

The final slide should be a recommendation, an ask, and the condition that would change your mind. Analysts often resist this on the grounds that the decision is not theirs. But withholding a view does not preserve neutrality; it simply hands the interpretation to whoever speaks next, who may have read the data less carefully than you.

Rehearse the objection, not the presentation

Before any significant presentation, write down the strongest argument against your conclusion and prepare an honest answer. If the honest answer is that the objection is partly right, say so in the room.

Audiences trust analysts who can articulate the case against their own recommendation. It is the clearest available signal that the work was done to find out rather than to persuade.