# Presenting research findings that lead to action

> A research finding only matters if someone decides something with it. Learn to structure a presentation around decisions and concrete evidence.

*Tags: ux, research, communication, stakeholders, decisions*

---


> [!info] Quick Definition
> Presenting research findings that lead to action means communicating them so that someone decides something based on them. A finding that is understood but changes nothing has not done its job. The goal of the presentation is to turn evidence into decisions, with owners and dates.


## Why is it hard for a finding to change anything?

Because researchers tend to present what they learned, in the order they learned it, while the audience wants to know what to do. That difference in starting point means many presentations are accurate and still produce no change.

Respondent warns against focusing on what users said they wanted and recommends turning research into actionable insights. That idea sums up the shift: you are not presenting data, you are presenting a proposed decision backed by data.

## What structure should you use?

Separate evidence from the insight that follows from it. Evidence is observation: what the user said or did. An insight is the actionable conclusion you build from those observations. A practical approach is to state up front what decision you need, explain the method without dwelling on it, and then show each insight with its evidence.

On the method, Baymard recommends explaining it in the first two minutes, before the recommendations, to pre-empt objections. It does not need to be long: what you did, with whom, and what its scope is.

For each problem, Baymard suggests pairing it with:

1. **A suggested fix.**
2. **The business consequence**, translated into what each person cares about: legal risk, cost efficiency, support calls, or customer satisfaction.
3. **The cost of inaction**, meaning the impact if the decision is not made.

Baymard also cites research by Anja Feibel Meerwald: stakeholders were more likely to act when the data was framed as "20% of users struggled" than as "80% of users had no issues". A negative frame can move people more, but use it with real data.

The close should be a list of next steps with priority, business consequence, effort, and timeline. Baymard also asks you to restate the permissions you need before finishing: know whether you are asking for money, resources, or approval. Without that list, the presentation ends in a conversation with no consequences.

## How do you use evidence?

An abstract finding ("users are confused by payment") persuades less than a concrete moment. Nielsen Norman Group presents video as a strong UX storytelling tool: it improves comprehension, builds empathy, and helps overcome skepticism. A clip of someone who cannot find the button shows the problem without anyone having to argue it.

Baymard also recommends relatable anecdotes and quotes from user surveys to humanize problems and make them concrete. Respondent suggests captivating the team with a story, such as following one user to show how they engage with the product, why, and what it means for the business, and then summarizing with metrics. Illustrative example: instead of saying only "30% abandon", tell the story of someone who reached payment in a hurry and left over an unexpected charge, and keep the number as backup.

Respondent adds that the data you show should match what your audience cares about, whether that is engagement, sales, or competition. And Baymard suggests replacing "we think" with "the data indicates", and avoiding hedging, which undermines even highly competent UX leads.

## How do you adapt it to the audience and handle skepticism?

Baymard suggests framing findings through each stakeholder's priorities, with examples such as legal risk, cost efficiency, support call volume, or customer satisfaction. Respondent recommends choosing the format based on how the team prefers to receive the message, from advanced visuals to a spreadsheet.

The same research can fit in a single document with three layers: a summary, a findings section, and an appendix of evidence. That is a practical suggestion, not a rule from the sources.

Baymard lists the most common objections: "we don't have data to prove this", "our users are different", and "there are bigger priorities right now". For the second one, it recommends relying on third-party research on standard web patterns. When doubt comes from opinion, offer external, measurable evidence, or even an A/B test on the site itself. Baymard also recommends neutral language: "the research indicates this user behavior" rather than criticizing the design.

## What happens after the presentation?

The presentation does not end when the last slide closes. A short email with the agreed decisions, owners, and dates prevents them from being lost. Baymard recommends keeping leaders updated on progress and building a culture of continuous measurement, and involving stakeholders in projects so they feel ownership.

If the finding turns into a change, show the result again. That closes the loop and makes it easier for the next study to get support.

## When should you NOT present?

If the study does not answer a concrete question, presenting it only adds noise. If the decision has already been made, the presentation becomes a formality. And if the team cannot act on the findings, it is better to agree first on what can be done before communicating them.

## Example

Illustrative example: a team presents to leadership the results of five interviews about the sign-up flow of a banking app. It opens with the decision it needs: simplify identity verification into two steps. In the first two minutes it explains how the study was done. For each problem it shows a suggested fix, the business consequence (drop-off during sign-up), the cost of inaction, and a 30-second clip of a person who does not understand why a document is requested. The last slide proposes three next steps, each with an owner, a timeline, and the permissions needed. Two weeks later, the team presents the results of testing the new flow. The conversation changed because the presentation asked for a concrete decision.

## Common mistakes

- **Dwelling on the method** and leaving the recommendation for the end.
- **Presenting findings without a recommendation**, which forces the audience to guess what to do.
- **Using only a quote or only a number**, without combining story and evidence.
- **Not preparing objections** before the meeting.
- **Hedging** ("we think") when data backs the finding.
- **Not following up**, and letting the decision dissolve.

To make findings findable later, [[atomic-research|atomic research]] proposes a way to store them as reusable units. And if the presentation proposes changing how the team works, [remote design collaboration](/en/wiki/processes/remote-design-collaboration/) explains how to share decisions when not everyone is in the same meeting. To understand the base method, [user interviews](/en/wiki/research/user-interviews/) describe how the evidence you later present is gathered.

## Sources

- [How to Present UX Research Findings to Get Buy-In From Stakeholders, Baymard Institute](https://baymard.com/blog/present-ux-research-findings-stakeholder-buy-in)
- [Seeing Is Believing: Using Video Evidence When Presenting Research to Stakeholders, Nielsen Norman Group](https://www.nngroup.com/videos/video-evidence/)
- [How To Present Your User Research Findings, Respondent](https://respondent.io/blog/how-to-present-user-research-findings)


---

Source: https://www.fernandoux.com/en/wiki/research/presenting-research-findings/
