Survey Results Presentation: A Six-Slide Worked Example

A survey results presentation should help the room decide what to do, not make everyone read a spreadsheet together. Start with the decision, show the evidence behind it, and keep the limits of the survey beside the numbers.

This guide follows a hypothetical customer-support survey with 120 valid respondents. You will turn its satisfaction ratings, multiple-choice answers and comments into six meeting slides. Once the totals are checked, you can turn the written findings into a presentation draft in Presenti; the statistical checks still happen before and after that step.

1. Define who answered and what the meeting must decide

For this example, the audience is a support team lead deciding which service improvement to investigate first. The survey covers recent support users who responded. It does not establish what every customer thinks.

Put the collection period, invitation method, eligibility rules and valid-response count on the opening slide. Record how you handled duplicates and incomplete submissions. Our teaching example assumes 120 valid respondents after those checks, but supplies no real collection dates or response rate. In your own deck, fill these in from the actual survey records, not from an AI-generated estimate.

A useful opening question is: “Which support improvement should we test next, and what additional evidence do we need?” That is more actionable than “Here are our survey results.” It also leaves room for the findings to be inconclusive.

2. Keep people, answers and selections separate

In the example, all 120 respondents answered the satisfaction question. There are 72 satisfied, 30 neutral and 18 dissatisfied respondents: 60%, 25% and 15%. These categories are mutually exclusive, so the counts sum to 120 and the percentages sum to 100%.

The improvement question allowed more than one answer. All 120 people answered it, but together they made 144 selections:

• Response speed: 66 people, or 55% of respondents.
• Clarity of explanations: 48 people, or 40% of respondents.
• Follow-up: 30 people, or 25% of respondents.

Here the percentages total 120%. That is not a calculation error: one person can appear in several categories. SurveyMonkey’s explanation of percentages above 100% describes this multiple-selection effect. Label the chart “Multiple answers allowed; base: 120 respondents.” Do not force the bars to sum to 100%.

Dividing by 144 instead would describe each option’s share of selections, not the share of people choosing it. Neither denominator can silently replace the other. Likewise, if only 110 people answer a later question, that question needs its own base.

There are also 45 optional written comments. A comment is useful evidence of an experience; it is not proof that all 120 respondents share it. Read and group the comments before reporting themes, note whether a comment can receive several theme labels, and keep the coding rules with the analysis. Do not invent theme counts or quotes. An anonymized excerpt can illustrate a theme, but it cannot establish its frequency.

3. Match the chart to the question

Use a simple ordered bar chart for the three satisfaction categories, or a 100% stacked bar when the distribution itself matters. Keep the neutral category visible. A high satisfaction share is not a reason to conceal dissatisfied responses.

For the improvement question, horizontal bars make the three options easy to compare. Show each count and percentage together, start the bar axis at zero and state the 120-person base. A pie chart is unsuitable here because the categories overlap; they are not slices of one whole.

Use the same color for the same meaning across slides. Highlight the option under discussion, not whichever bar happens to look attractive. If you compare two survey waves, first check that wording, audience, collection method and denominator are comparable. With only one wave, there is no measured trend to draw.

For open comments, show a small number of clearly defined themes and, where appropriate, anonymized excerpts. Keep the full question wording and detailed tables in an appendix. Removing clutter is different from hiding evidence that challenges your preferred conclusion.

4. Build six slides around the decision

1. Scope and meeting question. Identify the respondents, collection details and 120-person base. State the improvement decision the team needs to discuss.
2. Satisfaction distribution. Show 72 satisfied, 30 neutral and 18 dissatisfied. A defensible title is “72 of 120 respondents were satisfied with their support experience.”
3. Requested improvements. Show speed 66, clarity 48 and follow-up 30. Use “Response speed was the most selected improvement among these respondents,” not “Slow support is our biggest business problem.”
4. What the survey cannot explain. Separate the observed preference from possible causes. Ask whether queue time, repeated handoffs or unclear expectations might be involved; check operating records before choosing an explanation.
5. A proposed improvement test. Specify the change to try, a responsible owner, a measure and a review date. For example, investigate the handoff process before proposing a limited workflow test. Use real owners and dates once agreed; do not populate them with plausible guesses.
6. The decision needed today. Ask the team to approve an investigation or test, choose who will do it, and agree when to review the evidence. Keep unresolved questions visible.

The six-slide structure is a working outline, not a rule that every survey must fit. A short meeting may need only the decision, two findings and a proposed next step; detailed methods can sit in the appendix. A research audience may need more space for sampling and uncertainty.

5. Write conclusions that the data can support

“55% selected response speed” reports a preference in this survey. “Long waiting times caused dissatisfaction” claims a cause that these totals do not establish. Satisfaction and improvement choices have not even been cross-tabulated in this example.

A better headline is: “66 of 120 respondents selected response speed; investigate waiting and handoff times.” The first part is a result. The second is a proposed action. Keeping them distinct lets the audience challenge the action without losing the underlying fact.

For clarity, review anonymized examples of support explanations before proposing new wording. For speed, inspect relevant operational records before deciding that staffing or routing is the problem. After a change, compare like with like: an improved survey score alone does not prove the change caused it.

6. Give Presenti a checked brief, not raw responses

Use the Paste Text mode for a brief you have already prepared. Include the meeting purpose, exact counts, denominators, proposed slide order and claims to avoid. Remove names, contact details and individual case identifiers. An aggregated brief is enough for this example; there is no need to paste identifiable response rows.

You can adapt this input to your own checked results:

Create a six-slide presentation for a support team lead deciding which improvement to investigate. This is a hypothetical teaching example, not real customer evidence.

Base: 120 valid respondents. Satisfaction: 72 satisfied (60%), 30 neutral (25%), 18 dissatisfied (15%). Multiple-answer improvement question, base 120: response speed 66 (55%), clarity 48 (40%), follow-up 30 (25%). These are 144 selections, not 144 people. There are 45 optional comments; no quotes or theme counts have been supplied.

Slides: scope and meeting question; satisfaction; requested improvements; possible causes requiring investigation; proposed test; decision needed. Keep all counts and bases. Label multiple selection. Do not infer causation, invent comments, add a response rate or generalize to all customers. Leave actual collection dates, owners and review dates for us to fill in.

Review the draft against the checked source table. Inspect numbers inside charts as well as in text, including the 120-person base and the 45-comment limit. Check that a slide title has not strengthened “selected” into “caused” or “all customers want.” If you need to revise chart data later, confirm that the chart is editable in the output you intend to use; an image of a chart is not an editable data series. Keep the source table with the handoff.

The screenshot shows the actual input stage, not a generated or exported result. Presenti can help organize a slide draft; this workflow does not demonstrate automatic survey cleaning, statistical validation or causal analysis.

Questions that come up in the meeting

Do all survey questions need a slide?

No. Prioritize the questions needed for the meeting’s decision. Put full wording, remaining results and methods in the appendix, including findings that qualify the main argument.

Should I correct percentages that add up to 120%?

Not when multiple answers were allowed and the base is respondents. First verify the question type and denominator. In this example, 144 selections from 120 people correctly produce a total of 120% across the three options.

Can I call response speed the top priority?

You can call it the most selected improvement among these respondents. Choosing an operational priority also requires context such as severity, effort and service records. The survey informs that decision; it does not make it for you.