Writing an evaluation report

Use this page to learn about the process of writing an evaluation report.

Writing an evaluation report helps you share key findings and recommendations with those in your organisation and the people and communities you work with. This is the next step in the evaluation cycle after our guidance on analysing and reporting on your evaluation.

A report can be used to:

• suggest changes to how you work
• communicate your value to funders
• share good practice with other organisations
• share learning with the people and communities you work with.

Once you’ve completed these parts of your project, you’ll be able to write your evaluation report:

• You have data that you've collected and analysed.
• You’ve got the software to help you design your report.
• You have an understanding of the people who'll be reading your report.
• There are helpful colleagues available to read your drafts.

You have several options for software. Here are some suggestions below to get you started:

The Microsoft suite

• Word has a range of icons, images and smart art you can use - it is probably the most popular choice.
• Slide documents (using PowerPoint) can be helpful for writing briefer reports. You can also create data visualisation within PowerPoint and import it to Microsoft Word if preferred.
• You can create dashboards in Excel and/or import data visualisation graphs to other Microsoft applications.

Other applications

• SurveyMonkey has a dashboard function which can be used for reporting.
• Piktochart, Tablea and Canva are all design software. They have evaluation and impact report templates available.
• If you're producing content for webpages, Google Charts and Datawrapper may prove helpful.

Think about the people you're reporting to so you can tell them what they need to know. You should consider these points:

• What kind of information they need. For example, whether they need to know more about the difference you’ve made or the way in which you’ve delivered your work.
• How they'd like the information presented. For example, as a traditional evaluation report and/or data visualisation, webpages, or PowerPoint and when.
• Why they need the information and what you want them to do as a result.
• Whether there are any accessibility needs that you need to consider. For example, does the report need to work on a screen reader?

Having a clear structure makes your report easier to read. Before you write, plan your headings and subheadings. Most evaluation reports will include the following sections.

• Executive summary – a summary of your key findings and recommendations.
• Introduction – a brief description of what you're evaluating, the purpose of your evaluation and the methods you've used (for example, surveys and interviews).
• Findings and discussion – information on what you delivered, how you delivered it and what outcomes came out of it.
• Recommendations – actions that need to be taken to respond to the evaluation findings.

Reports will vary depending on the nature of your work, but you'll probably need to include findings on the following:

• Outcomes – What outcomes have been achieved, for whom and under what circumstances. You should also report on intended outcomes.
• Activities and outputs – What has been delivered, when and to who. You should also report on how satisfied the people and communities you work with were.
• Processes – Information about how you delivered your outputs. You may need this information to explain why something worked particularly well, or why it didn’t work.

Describe and interpret your data

In your report, you should describe your data and interpret it – analysing your data before you start writing will help with this.

Describing means presenting what the data tells you. You might describe, for example, what outcomes were achieved, by whom and in what circumstances.

Interpretation moves beyond description to say what the data means – make sure you word your report clearly so the reader can tell when you're describing data and when you're interpreting it.

To help you interpret data, you could do the following.

• Make connections by looking for trends, patterns and links. For example, if two groups had very different outcomes, what factors might have led to this?
• Put data in a meaningful context. Numbers don’t speak for themselves. Is 70% good or bad? How do you know?

When you interpret your data, you could discuss the following.

• Why outcomes were achieved, or not achieved. Understanding this may help you make decisions about future service planning. Many funders will also want to know about this.
• What worked and what didn’t. Knowing about this will put you in a good position to improve your work. It may also be useful to share with partners or funders to improve practice in the sector.
• Answers to your evaluation questions. When you planned your evaluation, you may have had two or three key questions you wanted it to answer. For example, you may have wanted to know whether your service works equally well for all groups.

Choose how to present your data

A common mistake is to try to present all your data, rather than focusing on what’s most important. It helps to narrow down to what people reading your report need to know.

It’s also important to think about how you'll present your information. You could consider the following points.

Which key numbers do your audience need to know?

• Decide whether to report using percentages, averages or other statistics.
• Think about whether you need to compare numerical data for different groups. You may want to look at whether men were more likely to experience outcomes than women, for instance.
• Read our guide on analysing quantitative data.

Which quotations will help you illustrate your themes?

• Choose quotations that bring your outcomes to life. Don’t choose too many or they'll distract the reader from the point you want to make.
• Have a mixture of typical responses and those that don’t fit easily into your categories.
• Read our guide on analysing qualitative data.

What visual aids will you use?

• Diagrams, graphs or charts should be used to highlight the most important information, rather than information which is less relevant.
• It’s very easy for diagrams to mislead your audience. Here are some examples of misleading charts. If you think a diagram might be misleading, it’s better to leave it out.

As far as possible, present data that has been analysed or summarised rather than raw data, to make it as easy as possible for the reader to follow.

Check anonymity and consent

When you collected your data, respondents will have said whether they wanted to remain anonymous (most do) and whether you should check with them before using a quote or case study in your report. Make sure you do any checking with plenty of time before you need to complete the report.

Depending on the size of your sample and how easy it is to identify individuals, you may have to do more than just change the name to make someone anonymous.

You might have to change their age or other identifying details, or remove references to anything that would allow people to identify them as an individual.

It’s important to write accurately and clearly so that your report can be easily understood and is not misleading.

Be transparent

Being transparent means being open about what you can and can’t say, and clear about how you reached your conclusions and about the limitations of your data. 

Just as it's important to minimise bias when collecting or analysing data, it's equally important to minimise bias when reporting.

• Avoid overclaiming your role in making a difference. Your work may not be solely responsible for the outcomes that have occurred for individuals or organisations you've worked with. Remember to report on evidence of any other contributing factors. For example, support received from other organisations or other sources.
• Choose case studies carefully. Evaluation case studies are not the same as marketing case studies. They should illustrate your learning points, not just the very best of what you do. You won't have a representative group of case studies, but as far as possible, choose case studies – and quotations – that reflect the full range of responses you had.
• Explore alternative interpretations or causal links. Sometimes, data is ambiguous and there could be more than one interpretation. All of us are prone to 'confirmation bias' – paying more attention to data that fits our existing beliefs. It's important to look for and talk about reasonable alternative interpretations or explanations of your data.
• Be clear about the limitations of your data. If there was a group you weren't able to hear from, or your sample over- or under-represents a particular group, say so.
• Be open about your sample size. In general, the smaller your sample, the less able you're to make generalisations about everyone in your target group.
• Report negative findings. If the data shows something isn't working or an outcome hasn't been achieved, don’t ignore it. Reporting negative findings will help your audience to use the evaluation to learn and improve.

Use precise language

Evaluation reports need to be as clear and precise as possible in their wording. Be especially careful about using the word 'proof' or 'prove'.

To prove something requires 100% certainty, which you are very unlikely to have. 'Indicates', 'demonstrates', 'shows', 'suggests' or 'is evidence for' are useful alternative phrases.

Make your report easy to read

Subheadings will make your report clear for your readers. Looking back at your evaluation framework or theory of change can help you think of ideas for subheadings.

It often makes sense to have a subheading for each intended outcome.

Sometimes you'll have collected data about the same outcome from a range of different sources such as questionnaires, interviews, observation or secondary data.

When you analysed your data, you probably looked at each source separately.

In your report, it usually makes sense to write about all the data relating to each outcome together (rather than having separate sections on data from different sources).

Keep your language simple and straightforward. Remember to explain any terminology that might be unfamiliar to your audience.

Your recommendations are likely to be one of the most important parts of your report. Good recommendations will make your evaluation findings more likely to be used.

Recommendations are more likely to be put in place if the following factors are considered.

• Supported by evidence – Be clear about how the recommendations build on the key findings. It can help to structure the recommendations in the same order as the main findings to help readers understand the evidence base for each.
• Specific – Say exactly what action needs to be taken and when within the control of the evaluation.
• Users – Make sure individuals or groups have the authority and capability to take forward what you’re suggesting.
• Realistic and achievable – Recommendations should be feasible. You can categorise them by which ones are easy to implement and which are less so. More ‘difficult’ recommendations might need budget or staff changes. These should still be stated, as well as the impact of it.
• Prioritised – It’s helpful to show some priorities for action. You could, for example, split your recommendations into ‘essential’ versus ‘optional’ or ‘for consideration’ versus ‘for action’. Make sure the number of recommendations you include is achievable.

You can involve other internal staff and the poeple and communities you work with at several points. For example, you could share your report drafts and ask them to help you refine the conclusions.

This 'co-production' of findings can be valuable and provide interpretations you may not have thought about.

You can also co-produce recommendations by sharing the findings with those you work with and asking them to suggest and prioritise recommendations.

If you do this, take care to guide people to base their recommendations on the evidence, and not their own interests or preoccupations.

Allow time for a couple of report drafts and make sure there are people available to review the report for you. It's good to have someone look at it with ‘fresh eyes’.

If the report is being widely shared, you could have someone from outside your sector review the draft to make sure it's clear for external audiences.

To complete the report, leave time for proofreading and editing, checking references, and design and print if needed.

You might include your data collection tools in appendices – this could help other organisations working in your field to improve their evaluation.

Once you’ve completed your report, read our guidance on using your findings to improve your work.