Picture the actual moment your report is read. It is not a quiet desk and a full hour. It is a busy person, minutes before a meeting, deciding in the first paragraph whether to keep going or to skim to the recommendations and wing it. For most of the people whose decision you are trying to inform, the report is the first half page. Everything after it is optional, and much of it is never opened.
This is not a failure of readers. It is the rational behaviour of people with more to read than time to read it, and it does not change because your analysis was careful. If the top of the report does not carry the finding on its own, the finding does not travel; it stays back in section four with the methodology, admired by nobody, moving nothing.
Most of a report exists to be checkable. The executive summary exists to be read. Confusing the two, writing the top as a preview of the reading instead of a replacement for it, is the most common way good evidence fails to land.
01 · Write it to stand aloneThe summary is the product, not the trailer
The test of an executive summary is brutal and clarifying: if the reader read only this and nothing else, would they know what you found, what it means, and what to do? Not "would they be enticed to read on." A trailer withholds to create suspense. A summary withholds nothing, because the person who needs it most is precisely the one who will not read on.
That means the summary has to contain the answer, not point at where the answer lives. "This report examines three options for the intake process" is a table of contents sentence; it tells the reader a topic and makes them go find the finding. "Of three options, only the second reduces the wait without adding staff; we recommend it" is a summary sentence; it hands them the finding and the recommendation in one breath. Write the second kind.
02 · What the first sixty seconds must carryFive things, in order
A reader who gives you a minute can absorb about five things. Spend them deliberately, and spend the first one on the answer to the question you were actually asked.
The top of the report, most important first
The widest band is read by everyone; each band below it by fewer people. Put the load-bearing sentence at the top, not the bottom.
The answer. Lead with the response to the commissioning question, in plain words. If they asked where the next dollar should go, the first sentence says where. The finding that carries it. The single most important piece of evidence, stated so a non-specialist gets it; one number or one pattern, not ten. The recommendation. What you think should happen, named clearly enough to act on. The biggest limit. The one caveat an honest reader must hold; naming it yourself is what makes the rest trustworthy. Where to look for more. A pointer into the body for the reader who does want to check the work.
03 · Say it the way they would say itPlain words are not dumbing down
The summary is the one place in the report where a term the reader would have to look up is a defect, not a precision. Confidence intervals, attribution language, and method names belong in the body, where a careful reader can weigh them. In the summary they do one thing: they make a busy person feel the report was written for someone else, and put it down.
A quasi-experimental design yielded a statistically significant between-groups differential (p < .05) on the primary outcome measure, subject to attrition-related threats to internal validity.
Participants improved more than the comparison group, by a margin unlikely to be chance. Because some people left partway through, treat the size of the gain as roughly right, not exact.
The second version loses nothing a decision-maker needs and gains the reader. Plain language is harder to write, not easier; it takes a full grasp of the finding to say it simply. The jargon version is often where an author hides the fact that they have not yet decided what they actually found.
If you cannot say it in the words your reader would use, you may not yet understand it well enough to summarize. The struggle to simplify is the last mile of the analysis, not a betrayal of it.
04 · Grade your own topA sixty-second checklist
Before a report leaves your hands, read only its first half page and run it against the six checks below. If your top cannot pass them, the reader who never gets past it will not get what they came for. Tick what your summary already does; the readout keeps score as you go.
The sixty-second checklist
Read only the first half page of your report, then tick each one that is true of it. Nothing is uploaded or kept.
Tick each check your summary already passes.
Aim to pass all six before the report leaves your hands; the reader who never gets past the top is the one you most need to reach.
05 · The counterargumentA summary can oversimplify into a lie
Pushed too far, "make it simple" becomes "make it wrong." A finding stripped of its one essential caveat is not usable evidence; it is a confident sentence that will mislead exactly the person who trusted the top and skipped the body. Simplicity that drops the load-bearing limit has not served the reader; it has set them up.
The discipline that keeps a summary honest is the fourth check: the biggest limit stays, even at the top, even in plain words. A good summary is short because it is precise, not because it is incomplete. It says the true thing in fewer words; it does not say a smaller, easier, slightly false thing. If simplifying forces you to drop the caveat that changes what a reader should do, you have found the edge, and the caveat wins.
Recommendation: write the summary first, before the body, as a forcing function. If you cannot yet write a clear top, you do not yet know what you found, and the analysis is not finished no matter how many tables it has. Then write it again last, once the body is done, and cut it until every sentence in it is load-bearing. The top is the last thing you polish and the first thing anyone reads; treat it accordingly.