How SWR brings its market research into everyday campaign work.
SWR's market research differentiates nine personas. In day-to-day work they were barely usable: study results are static, and a standing panel is too expensive. RSN8 operationalized the nine personas as AI agents. Now the team tests campaign material against them directly.
„RSN8 takes A/B testing to another level. Getting direct feedback without complex surveys provides more confidence in campaign development.“
Starting point
Nine personas that only existed on paper.
Südwestrundfunk (SWR) is the public broadcaster for Baden-Württemberg and Rhineland-Palatinate, reaching millions of people via TV, radio, podcasts, and the ARD Mediathek. For the campaign of the Instagram channel “SWR Heimat”, the marketing team faced the question: which stories and messages connect the audience most strongly with the region?
SWR's market research provides a differentiated basis: nine personas with similar shares of the population. But the static study results were hard to embed in the running campaign process, and a standing panel for every question would be too expensive. Add the similarity-attraction effect: marketing teams, too, unconsciously pick content based on their own preferences rather than the audience's.
Approach
Study results become AI agents.
RSN8 aggregated SWR's market-research results and modeled the nine personas as digital AI agents. Campaign narratives, social-ad drafts, and messaging variants can since be tested against each individual persona in a structured way, with direct feedback and concrete optimization suggestions.
Market research before
- Nine personas as a study report in a folder
- Follow-up questions only with the next survey
- Barely tangible in the running campaign process
With RSN8
- Nine personas as AI agents, available any time
- Every draft testable against all personas directly
- Feedback and follow-ups within minutes
Result
Faster validation, tighter decisions.
What counts for SWR is that market research now arrives inside the running process. With the AI-backed recommendations, the team could:
- Validate campaign messages faster, without costly live focus groups
- Tailor social ads more precisely to the needs of each target group
- Tighten decisions, because audience feedback replaces subjective debates
The simulated reactions of the nine personas guide creative selection and reduce the influence of the similarity-attraction effect.