How to Structure a Research Debrief Presentation (October 2026)

A research debrief presentation works when it is built backwards from the decision, not forwards from the report. You open with the choice the audience faces, state your answer with its confidence level in the first three slides, then build the evidence one idea at a time so nobody has to guess where the argument is going. The rest of the session follows the same logic: findings, limits, recommendations, decisions with owners.

Most debriefs fail in a predictable way. Someone takes the written report, cuts it into pages, and reads it aloud. That deck is too heavy to present live and too thin to forward afterwards, so it does neither job well. Knowing how to structure a research debrief presentation is what separates the two.

Here is the short version before the detail:

• Start with the decision at stake. Name the choice, the person who owns it and the date it has to be made.
• Give the answer by slide three. The headline finding and its confidence level appear before any methodology.
• Sequence slides as a chain of questions. Each slide answers one question the previous one raised.
• Show what the research could not establish. One slide, in the main flow, never a footnote.
• End on decisions, owners and dates. A readout with no decision taken was a status update.
• Move everything else to the appendix. Full tables, instrument, quotas and code frames live there.

Those six lines are the spine. The rest of this guide fills in the mechanics: what to prepare, how to write each slide, how many slides the session can actually hold, and what to do when a stakeholder challenges a finding in the room.

What You Need

Most of what makes a debrief feel focused was decided days before the room, in preparation, not in the room itself.

The decision at stake, in one sentence

If you cannot write the decision in a single sentence, you are not ready to build the deck. Something like “should we move the trial packaging from the current 500g format to two 250g packs for the second half of the year” tells you what evidence matters. “Findings from the packaging study” does not.

Write it, then name the person who owns that decision and the date by which it must be made. If nobody in the room can own the decision, you are presenting an update, not running a debrief, and the structure changes accordingly.

The audience list, with a role against each name

Sort everyone into three groups: decision-makers who need to act, contributors who supply context or expertise, and observers who need awareness only. Mixed audiences are where decks go wrong, because a designer and a finance lead need different evidence from the same study.

For a mixed room, put the commercial and behavioural evidence in the main flow and push the technical detail to the appendix. If someone in the second group cannot follow the main story without help, the story is pitched wrong, not the person.

The questions the deck has to answer

Write down three to five questions in the words your stakeholders actually use, not the words the report uses. “Will people pay more for less packaging” is a real question. “Do respondents associate refill with sustainability” is a research objective, and it belongs on the method slide rather than the argument.

Every slide you keep should trace back to one of those questions. Every slide that does not is a candidate for the appendix.

An agreed evidence standard

Decide in advance what the study can support and what it cannot. For quantitative work that means the base size for each cut, the confidence level you are prepared to stand behind and the minimum subgroup size you will quote. For qualitative work it means how many participants said something before you will call it a theme.

Write these thresholds down before you read your own charts. Once you are deep in the data, it is easy to quietly relax a rule because an interesting cut appears.

A send-ahead pack and an appendix folder

Send the deck, or at least the executive summary and the agenda, 48 hours ahead. People who walk into a readout cold spend the first fifteen minutes forming an opinion about the study, and they rarely revise it.

Prepare a separate appendix folder with full tables, the instrument or interview guide, quotas, the coding frame and any wave-over-wave detail. Number the findings (F1, F2, F3) and the recommendations (R1, R2, R3) so that people can reference them in the discussion and in the follow-up email.

A time budget, a facilitator and a note taker

Reserve a third of the session for discussion and protect it. The most common failure in the room is not a hostile question, it is the presenter running eleven minutes past the point where the decision needed to be made.

Use a facilitator who is not the researcher when the findings are contested or the stakes are high. The person presenting the evidence cannot also police the conversation, and the second role usually wins over the first.

Step-by-Step: How to Structure a Research Debrief Presentation

Structure a research debrief presentation in eight moves: frame the decision, give the audience a three-part storyline, establish enough method to build confidence, present findings in a repeating evidence pattern, synthesise them into themes and tensions, translate those into implications, turn the implications into testable recommendations, then close on decisions, owners and next steps. Each move has a clear output and a simple check that it worked.

1. Start with the decision, not the methodology

The first slide is not “objectives” and it is not a logo. It states the decision the room faces, the person who owns it, the date and the cost of getting it wrong in either direction.

Slide one, roughly: the decision, the owner, the date. Slide two: the answer in one sentence, with the confidence level attached. Slide three: the three-part storyline for the session, so people know how long it runs and where their questions belong.

The test: read only the first three slides to someone who was not in the room. If they can describe both the decision and your recommendation, the opening works. If they cannot, you are still writing background.

2. Give the audience a short executive storyline

Most decks fail because nobody in the room knows how many findings there are or how the pieces connect. A three-part storyline fixes that in about fifteen seconds.

• Where we are. One slide of context: the market, the customer question, the business problem. No history.
• What we found. The central discovery, usually three to five supporting findings that each explain part of it.
• What it means for us. The implication and the recommendation, including what you would not do.

Give each part a plain signpost and one sentence that sums it up. “Three things we were wrong about” is a better signpost than “Section 2: Findings”, because it tells a sceptical executive exactly what kind of slide is coming.

Here is the whole deck on one page. The example headlines come from a fictional packaging study, but the structure transfers to any project:

Slide	Purpose	Headline says	Evidence on the slide	Time
1	The decision at stake	The choice, its owner and the date	None. Context only	1 min
2	The answer	What we recommend, with confidence	None	1 min
3	Storyline	Three parts, with time per part	None	1 min
4	Context	What the business is trying to solve	One chart or one image	2 min
5	Method, in brief	Sample, recruitment, fieldwork dates	Base sizes and quotas	2 min
6	Central finding	The discovery in one sentence	One dominant chart	3 min
7	Finding F1	The finding as an assertion	Chart plus two quotes	3 min
8	Finding F2	The finding as an assertion	Chart or concept map	3 min
9	Finding F3	The finding as an assertion	Chart plus segment split	3 min
10	Tensions	Where the evidence contradicts itself	Two columns side by side	2 min
11	Limits	What this research could not establish	List of gaps	2 min
12	Implications	What we should consider doing	Three boxes	2 min
13	Recommendations	R1, R2, R3 with confidence and owner	Table with owner and date	3 min
14	Decisions and next steps	What is decided, by whom, by when	Action list	3 min

Notice what is missing: an agenda slide, a thank-you slide, a team credits slide inside the main flow, and every individual chart. That is the point. Credits belong on a closing frame or in the appendix, and the individual charts are already inside the findings.

3. Establish the research context and approach

The method slide is not an apology and it is not a lecture. It exists so the audience can calibrate how much weight the findings can take, which means four things and nothing else: the objective in one line, who was included and how many, when fieldwork ran, and the limitations that matter for the decision at hand.

Put base sizes on the slide, not in the appendix. A percentage without its base is a decoration. “62% of pack buyers said they would try refill” means something quite different from “62% of 18 to 34s in urban households said they would try refill”, and the audience cannot tell which one they are looking at unless you say so.

Skip anything that does not change how much someone should believe you. Recruitment channel matters when it affects who answered. Screening criteria matter when a subgroup is central. A list of every question in the questionnaire does not.

The check: a sceptic should be able to tell from this slide alone where your evidence is strong and where it is thin. If they cannot, add the missing detail here rather than waiting for the appendix.

4. Present findings in a consistent evidence pattern

Every finding slide should follow the same four-part pattern, so the audience learns the shape once and stops spending attention on your slide design.

1. The finding, stated as a full sentence headline.
2. The evidence: one dominant chart, or one chart and two supporting quotes.
3. What it means, in one line.
4. What we should consider doing about it.

That last line is what most decks omit, and it is what turns a finding into something a stakeholder can act on. “Refill interest is highest in the 25 to 34 group” is an observation. “Refill is worth testing as a subscription mechanic for that group rather than a one-off purchase” is an implication.

Write headlines as assertions, not labels. “Segment results” tells the audience nothing. “Older buyers are not interested in refill at any price point” tells them the finding before the chart loads, and if the chart contradicts the headline they can stop you immediately, which is exactly what you want.

Keep one dominant element per slide. Two charts side by side produce a conversation about the charts rather than about the finding, and the finding is the thing you are there to deliver.

For qualitative evidence, treat quotes as evidence rather than decoration. Anonymise them, mark the base they come from and keep them short. A single quote illustrating a theme you established across a coded set is fine; a quote standing in for that evidence is not.

5. Connect findings to audience needs and tensions

Individual findings are not a synthesis. The synthesis slide is where you group them into themes, needs, barriers and segments, and where you show the contradictions.

Three patterns recur in most studies and each needs a decision. Unanimity means the finding is solid and the argument is about execution, which is the easy case. Segmentation means the finding is real for one group and absent for another, and the debate is about priority. Tension means two findings point in opposite directions, and you need to explain why before anyone will act on either.

Show tensions explicitly rather than hoping the sharper slide wins. Two columns, the contradictory finding on each side, and your read of what is driving it. Practitioners who have run these sessions report that raising the contradiction first is what keeps a sceptical stakeholder from spending the meeting trying to find it.

The check: after this slide, the central finding should look inevitable rather than merely stated. If your synthesis is a list of themes with no hierarchy, you have not synthesised yet.

6. Translate insight into implications

Implications are the bridge between what the research says and what the organisation might do, and they are where most decks jump straight to recommendations and lose credibility.

Split them into three columns on one slide: opportunities, risks, and open questions that need more work. The third column is not padding. It tells the audience which parts of the story you are not claiming, and it converts “you have not thought about X” from an attack into something you already flagged.

Keep implications conditional where they are conditional. “If refill interest survives contact with shelf prices” is a better line than “adopt refill”, because it tells the room exactly which assumption the recommendation rests on, and you can point back to it when that assumption is challenged.

The check: for each implication, can you name the finding it came from? If an implication has no finding behind it, it is an opinion wearing a research finding’s clothes, and someone will take it apart.

7. Make recommendations specific and testable

A recommendation is only finished when someone could act on it without asking a follow-up question. For each one, name the audience or behaviour it targets, the rationale that connects it to a numbered finding, a confidence marker, a named owner and a next step.

“Explore refill further” fails on every count. “Test refill packs as a subscription option with urban 25 to 34s, based on F1 and F3, medium confidence, owner Sam in Category, next step is a concept test in six weeks” passes.

Confidence markers are worth the effort even when they are uncomfortable. High for something the data supports directly, medium for a supported inference, low for a hypothesis worth testing. Marking a recommendation low confidence tells the room it is a proposed experiment rather than a conclusion, and that framing changes how people respond to it.

Rank by decision value, not by certainty. The lowest-confidence recommendation might still be the most important one to fund this quarter, and the deck should say so out loud.

8. Close with decisions, owners and next steps

The final slide is an action list, not a summary. Each row: the decision, who owns it, the date, and the sign-off status. If nothing was decided in the room, write that down as well, because “no decision, revisit on 14 November” is a legitimate and far more useful outcome than leaving the room with an agreement in the hallway.

Hold the discussion to four questions, and put them on the agenda slide so nobody has to invent them:

1. What has to be true for us to act on this?
2. Which finding are we least comfortable with, and why?
3. What would change the recommendation?
4. Who owns each next step, and by when?

Answer the first two before you reach the appendix. The rest can run as long as the room wants, but keep the evidence behind you available on screen rather than behind a hidden folder.

Now the timing. The rule of thumb from experienced facilitators is two to three minutes per substantive slide, with a third of the session reserved for discussion. That arithmetic is worth writing on a sticky note before you build the deck, because it decides how many slides you can afford.

Session length	Presentation time	Discussion time	Slide budget
30 minutes	18 minutes	12 minutes	8 to 10 slides
45 minutes	30 minutes	15 minutes	13 to 15 slides
60 minutes	38 minutes	22 minutes	16 to 18 slides
90 minutes	55 minutes	35 minutes	22 to 25 slides

If a forwarded deck also has to work read-alone, you have a second artefact to think about. The live deck carries the argument in its slide order; the read-along deck carries it in speaker notes and captions, because a person reading it alone cannot hear your caveats. The practical fix is to add short captions under each chart that state the finding and its base, and to write speaker notes as if the reader is a sceptic who missed the meeting.

Common Mistakes

Each of these has a direct fix, and the fix is usually structural rather than cosmetic.

Opening on objectives

The audience does not need to know how the study was framed before they know what it found. Move objectives into the method slide at slide five and use the opening for the decision and the answer.

Reporting every chart

A deck that contains every exhibit in the report is a report with page breaks. Choose one dominant chart per finding and move the rest to the appendix, ordered by the question someone is most likely to ask.

Dense slides that get read aloud

If a slide needs a paragraph of explanation, it is carrying two ideas or the wrong idea. Cut it, split it, or move the text into speaker notes. Whitespace is not wasted space when the reader has to absorb an argument at speed.

Confusing insight with implication

An insight describes what you learned. An implication describes what follows for the organisation. Label them differently on the slide, because the second is what people will argue about and it needs to be visibly distinct from the evidence.

Hiding the conclusion and the limitations

If the answer sits on slide twenty and the gaps sit in a footnote, nobody has the full picture when they start challenging. Put the limits in the main flow, on their own slide, with a headline that states them as a finding about the evidence rather than a caveat.

Ending without a decision

A readout that ends with “we will circulate the full report” has produced nothing. Build the closing action list backwards: for each open decision, name the owner and the date before the session and ask for a signature in the room.

Running long because discussion was never reserved

Reserve a third of the session in the invitation, not just in your own plan. If the room overruns, cut the deck rather than the discussion block, and use the order below to decide what goes.

Letting pushback arrive unprepared

Assume three to five questions: how strong is the evidence, how representative was the sample, why does this contradict the last study, why are we not doing nothing, and who is accountable for the next step. Each answer already exists somewhere in the deck. Put the support in the appendix in the order you expect the questions, so the answer is one click away rather than improvised.

The cut order for an overlong deck

Cut whole slides before you cut inside a slide, and never cut base sizes or confidence markers to save space. Then drop supporting findings that do not change the recommendation, then detail that belongs in the appendix anyway. Anything you cannot defend in one sentence as part of the argument does not survive the cut.

Frequently Asked Questions

Conclusion

Start with the first action, not the software: before you open a slide template, write one sentence for the decision at stake, one for the answer with its confidence level, and one for the outcome you want from the session. Everything after that is sequence, and sequence you can edit in an afternoon.

Build the deck backwards from the closing action list, keep one idea per slide with the finding written in the headline, and give the limitations their own slide. When the deck is forwarded a week later to someone who missed the meeting, those three habits are the reason it still makes sense.