A media organisation can know, down to the minute, how long people spend on an article. It knows how many started a video, when they stopped watching and how many shared an excerpt. It knows much less about what those people understood.

It is a strange blind spot. We have increasingly precise tools for tracking the circulation of content, yet its reception often remains an assumption. Two readers can leave the same report with opposing interpretations. A crucial piece of information may be remembered, forgotten, discussed with someone close or rejected because it clashes with a longstanding conviction. Distribution can be counted. What happens in people’s minds and lives takes more work.

Yet this is where a media organisation’s impact begins.

Research on media effects has long abandoned the idea that a message produces the same response in everyone. We choose some of the information we encounter. We approach it with a history, knowledge, loyalties and sometimes well-established fatigue or mistrust. We do not all remember the same details. A review published in the Annual Review of Psychology describes effects that vary with people, content and the conditions in which they meet. Audiences are not the final stage of a production chain. They help shape the meaning of what we publish.

This should matter to every newsroom, including those that do not describe their work as “impact journalism”. A journalist chooses what merits an investigation, the facts that open a story, the voices we hear and those still being sought. A newsroom decides to follow a subject for weeks or move on. These decisions affect what becomes visible and how a problem can be understood.

McCombs and Shaw’s work on agenda-setting showed the importance of the prominence media give to issues. Robert Entman’s work on framing helped describe what our presentation choices foreground or leave in the background. This research does not say journalists can dictate citizens’ opinions. It makes a still-convenient professional belief harder to sustain: once the facts are published, our work is done and what follows is no longer our concern.

I do not believe in that separation.

No publication is neutral in the sense of involving no editorial choices and having no possible effects. That does not mean every article changes an opinion or triggers a measurable action. A report may reinforce a conviction, raise a doubt, draw attention to a neglected actor or provoke rejection. It may also leave no trace we know how to observe. Precisely because the effect is neither uniform nor guaranteed, it is worth trying to understand.

At OurFrame, we use Theory of Change to organise this inquiry. The method begins with a question too often asked at the end: what do we hope will change, and for whom? It then requires us to describe the assumed steps between content and that effect. The people concerned still need to encounter it, pay attention, understand what it offers and connect that understanding to their own situation. Something may happen at each step. At each step, the hypothesis may also fail.

This way of working changes the early discussions. “Informing the general public” is no longer enough. Which public, exactly? Are they already familiar with the subject, or discovering it? What might they misunderstand? What story could help them grasp a mechanism rather than remember an isolated fact? And if we hope for action, is that action really within their control? Someone can understand a problem perfectly and have no means to respond. Asking them for more motivation then amounts to the wrong diagnosis.

Theory of Change helps reveal these gaps before they disappear into a distribution report. It does not turn an intuition into scientific certainty. It writes it clearly enough to be discussed, tested and, if necessary, abandoned. Our methodological framework distinguishes the content produced, exposure to it, initial audience changes, intentions and behaviours. It also asks us to identify obstacles and other possible causes of an observed change.

This standard matters particularly in investigative journalism. A revelation may be picked up by other media, support a mobilisation or prompt an institution to open an inquiry. Publication activates none of these mechanisms on its own. Sometimes it takes people who bring the facts into public debate, organisations able to act on them and officials required to respond. The Global Investigative Journalism Network documents both the variety of these effects and the difficulty of attributing them to a specific investigation. Following what happens after publication does not diminish journalism’s value. It helps us understand its actual reach, including when that differs from what we imagined.

Media organisations readily discuss their influence. They claim it when an investigation prompts a response; they fear it when the discussion turns to misinformation, polarisation or manipulation. Between these attitudes, a more patient practice is often missing: studying how our content is received, by whom, under what circumstances and with what effects.

We have learned to examine our audience figures very closely. It is time to give the people behind them the same attention. Perhaps the most useful question after publication is not “how many saw it?”, but “what did they see in what we thought we were showing them?”