Your community satisfaction survey gives you one headline number, and everyone in the room reads it differently. Some see a pass mark, some see a problem, and few can say what it means or what to do next. This post shows how to read that score against the right benchmarks, find the services driving it, and set up a way to track change through the year. The aim is simple: leave the meeting with a plan, not just a slide.
What the 2026 results say
The 2026 Victorian results rolled out through winter, council by council. Local Government Victoria puts the statewide average for overall council performance at 58 out of 100. That figure is a benchmark, not a target. It tells you where the middle of the pack sits this year, so you can see whether your own result is above it, below it, or close to the line.
- 58 out of 100
- Statewide overall performance benchmark
- 64 out of 100
- Metropolitan average overall performance
- 400, selected at random
- Minimum residents surveyed per municipality
- 70 out of 100
- Example single-council result (City of Boroondara)
Read those numbers together and the picture changes. A score of 58 is the statewide average. Metropolitan councils sit higher on average, at 64 out of 100. The City of Boroondara reported 70 out of 100. One number on its own tells you very little until you know which group you belong to and what a fair comparison looks like.
A single score hides the story
An overall score is a summary of many smaller judgements. Underneath it sit category ratings across the things your council does: roads, waste, planning, customer contact and community consultation. Residents rate each area, and the headline rolls them up. Two councils can share the same overall score for very different reasons, so the useful work starts one level down.
The sample matters too. The survey is run by the Department of Government Services, with a minimum of 400 residents and ratepayers aged 18 and over per municipality, selected at random. That is enough to trust the overall direction, but small groups within the sample can move from year to year. Before you react to a one point shift, check whether it is a real signal or ordinary noise.
