At what point does audience segmentation stop being rigour and start being a way to avoid deciding?
I’ve seen change plans with segmentation grids so granular they take longer to build than the campaign itself, and most of that detail changes nothing. A segment only earns its place if it clears three tests: does this group face a genuinely different impact from the change, would the same message land differently for them because of role, location, tenure, language or access, and can you actually reach them through a distinct channel or a credible voice. Fail any of those and the segment is a spreadsheet exercise dressed up as strategy.
You’ve segmented enough when another layer stops changing the message, the channel or the timing, and only changes a label on a report. For most large change programmes that’s five to eight real segments: frontline versus office-based, tenured versus new joiners, people leaders versus individual contributors, geography or language cluster.
Past that point, teams aren’t closing a comprehension or trust gap. They’re producing evidence of thoroughness for a stakeholder who’ll never read the grid.
How many of your current audience segments would anyone notice if you deleted them tomorrow?
