How can we find the story in the data?
Category: Blogs
26 August 2026
Every day we're flooded with data. Election results, economic indicators, climate records, breaking news. But how do you turn millions of data points into something people can understand, trust, and act on?
In the latest episode of Connecting Society, host Shayda Kashef-Tomlins meets Alan Smith OBE, Head of Visual and Data Journalism at the Financial Times, to explore the art and science of data storytelling.
Drawing on a career spanning official statistics at the Office for National Statistics (ONS) and journalism at the Financial Times, Alan discusses what makes a compelling data story, how visualisation can reveal what averages hide, and why communicating complex evidence clearly doesn't have to mean oversimplifying it.
Start with a question, not an answer
What separates data journalism from traditional journalism? According to Alan, it's not just about starting with data - rather, it's about starting with a question.
"Traditional journalism might start off with the sort of the hunch or the tip or the seed of a story idea," he says. "Instead, what we really mean when we talk about the purer forms of data journalism is where we're saying, well, actually, let's start with a question."
This approach is almost academic in nature. As FT data journalist John Burn-Murdoch described it, data journalism is “social science on a deadline”; testing hypotheses through original research before determining the story's top line.
Beyond averages
One of Alan's most memorable examples illustrates why moving beyond averages matters. A viral chart showed countries' healthcare spending per capita against life expectancy over time. Most countries followed an expected pattern, where more spending correlated with longer lives.
Except the US, which appeared as a dramatic outlier – spending far more but seeing no commensurate improvements in life expectancy.
Looking beneath the average reveals substantial inequalities in how healthcare spending is distributed. For Alan, this is where data storytelling becomes particularly valuable. Rather than stopping at the headline figure, it can help audiences explore the variation, inequalities and experiences hidden underneath it.
This is especially relevant when working with population-level data. Administrative data can allow researchers to move beyond national averages and investigate how experiences and outcomes differ between groups, places and over time.
Clarity doesn't have to mean simplicity
Communicating complex evidence to a wider audience often comes with pressure to simplify. Alan suggests a slightly different goal: prioritise clarity over simplicity.
Simplification can involve stripping information away. Clarification, by contrast, helps audiences navigate complexity by creating hierarchy and providing context.
A chart, for example, might contain ten lines rather than one, but visually draw attention to the line that matters most while retaining the others as useful context. The same principle applies to written communication: audiences can engage with nuance when information is structured in a way that helps them understand what matters and why.
It also means thinking carefully about who the audience is and how they are encountering the information. A visualisation accompanying breaking news may need to be understood at a glance, while someone reading an in-depth analysis may be willing to spend several minutes exploring a more detailed graphic.
Put the user first
Alan's final piece of advice for researchers working with population-level data is straightforward: make communication user-focused rather than producer-focused.
He recalls developing interactive population visualisations at ONS, where separate teams initially wanted separate products for population estimates and population projections. From an organisational perspective, that made sense because they were produced by different teams for different purposes. From the user's perspective, however, combining historical and projected population change into one experience was far more useful.
It's a simple example with a wider lesson. Whether communicating administrative data through a chart, report, website or story, start by asking what the audience needs to know – rather than how the information happens to be structured behind the scenes.
For researchers hoping their findings will reach policymakers, practitioners or the public, such a shift in perspective can make the difference between simply publishing evidence and communicating it in a way people can understand, remember and use.
Listen to the full episode now
Listen to the latest episode of Connecting Society to hear Alan and Shayda's full conversation on data journalism, visualisation, trust and what it takes to turn complex information into meaningful stories.
