Turning monitoring data into stories people remember

Most programmes collect far more data than they communicate. Indicator tables, baseline surveys and quarterly reports hold real evidence of change, but few people outside the M&E team read them. Data storytelling is the work of turning that evidence into findings people understand, remember and act on.

The United Nations Economic Commission for Europe (UNECE), whose guidance was written for statistical offices, puts it bluntly: without a story line, a release becomes just a simple description of numbers.¹ Its starting point is the audience, which it says is "in the driver's seat".¹ A minister, a donor and a district health committee need different findings from the same dataset.

Before anyone opens a design tool, write down the one message the data supports. The EU's data visualisation guide suggests pausing to think about the purpose of a visualisation and asking what message you are trying to communicate to your audience.² If the team cannot agree on that sentence, no chart will rescue the report.

A useful test is to complete the phrase "The data show that..." in under 20 words. "Enrolment rose" is a start. A hypothetical finding such as "More girls stayed in school after the second year, and the gap with boys has closed in three of four districts" is a finding. It says what changed, for whom and compared with what.

Write that sentence together. M&E staff know the caveats, the sample sizes and which results are robust; communications staff know the audiences and the channels. In our own practice, a short working session between the two, held before drafting starts, saves several rounds of corrections later and keeps the finding both accurate and readable.

Journalistic structure works for results. UNECE advises that the most critical information must come at the beginning, because readers lose interest quickly, and that the lead is the most important element of the text.¹ It recommends subheadings that summarise the main finding, ideally with a verb, and chart titles that state in a few words the story the chart illustrates.¹

So instead of a heading such as "Outcome indicator 2.1", write "More farmers now sell through cooperatives". Instead of a chart titled "Household income, 2022 to 2025", write a title that says what the line does. The indicator code can sit in a note for the specialists who need it.

Numbers persuade only when they are few and meaningful. UNECE advises against burdening readers with too many numbers, using only key rounded figures, and using the text for analysis, trends and context rather than repeating table values.¹ It also asks writers to place the most important findings in the context of short- and longer-term trends, while taking care not to lose too much information when rounding.¹

In practice, give every headline number a reference point: the baseline, the target, the national average or the same period last year. Write numbers as numerals; Nielsen Norman Group found that digits attract the eye of readers scanning a page.³ And translate percentages into people where you can, as "one in four households" is easier to picture than "24.6%".

Numbers show scale; people show meaning. A Catholic Relief Services and American Red Cross guide on writing human interest stories for M&E describes them as a way to personalise impact results and a complement to other data collection techniques.⁴ It distinguishes success stories, which illustrate impact through one person's positive experience, from learning stories, which draw lessons from positive and negative experiences and suit ongoing M&E reporting.⁴

The same guide advises focusing on qualitative information with supporting quantitative data, opening with who, what, where, when, why and how, including direct quotes, and keeping stories to around 500 to 750 words.⁴ Placed next to the headline finding, one well-chosen voice helps readers remember the number.

Credibility is the whole point of results communication. The CRS guide notes that individual stories do not have to be representative, but asks organisations to be transparent about how subjects were selected and to state the limits of generalising from them.⁴ It also warns against overdramatising, which can reduce credibility.⁴

Be equally careful with cause and effect. BetterEvaluation explains that outcomes usually result from a combination of factors, and that an intervention contributes to change without being the sole factor producing it.⁵ Write "contributed to" rather than "caused" unless an evaluation shows otherwise, and name partners and government systems that played a part.

Stories drawn from monitoring involve real people, often in vulnerable situations. The CRS guide calls for consent before interviewing, observing or photographing anyone, and for not sharing information people would prefer to keep private.⁴ It also observes that projects often spend time developing good stories but set aside too little time to share them, and recommends planning dissemination in advance.⁴

Plan the formats when you plan the data collection: a one-page brief for decision makers, a short video or photo story for the public, social media posts built around a single finding, and the full report for specialists. The same evidence, told well in each format, reaches far more people than a single long document.

• One-sentence message agreed before design starts
• Headings and chart titles that state findings
• Few rounded numbers, each with a baseline or comparison
• Indicator codes moved to notes
• A human story placed next to the key finding
• Story selection and limits explained
• "Contributed to" rather than "caused" unless evidence shows otherwise
• Consent recorded and dissemination planned