AI Can Help You Understand Your Audience. Here’s How
Clicks are easy to count. What people actually need is much harder to see. A story can perform well and still leave a newsroom guessing. Did people find it useful? What were they trying to understand? Why did they stay, leave or ignore it altogether? Khalil helps newsrooms answer those questions. His work brings together audience development, user needs and AI, with one aim: helping journalism understand the people it is supposed to serve. In this AImpactful conversation, Branislava Lovre speaks with Khalil A. Cassimally about what audience connection looks like beyond reach and engagement figures. They discuss the User Needs Model and what changes when a newsroom stops asking only what it wants to publish and starts asking what people need from it. Khalil explains how that shift can bring more clarity to editorial decisions and help teams create journalism that people find genuinely valuable. Quantitative data can show what people did. Qualitative research can help explain why. Khalil discusses how AI can help newsrooms analyse interviews, responses and other complex material, and identify patterns that might otherwise be missed. The conversation also looks at synthetic personas. Khalil explains where they can help teams test ideas and question assumptions, but also why they should support research with real people, never replace it. Branislava and Khalil then turn to the growing number of AI tools available to newsrooms. Khalil’s advice is simple: start with the problem you need to solve, then choose the tool. He also shares how he uses AskRally, ChatGPT, Julius.ai, n8n and Granola in his own work. What we explore: What audience connection means beyond clicks and reach How the User Needs Model changes newsroom decision-making How AI can help analyse qualitative audience research The difference between audience behaviour and audience motivation What synthetic personas are and how newsrooms can use them Why synthetic personas cannot replace conversations with real people The risks of stereotypes, incomplete data and false confidence How to choose an AI tool by starting with the problem Tools for research, data analysis, automation and note-taking Why better technology does not remove the need to listen