Why most manufacturing dashboards fail (and how to fix them)
Short answer: A dashboard fails when it answers questions nobody asked. Design backward from the decision.
I've seen beautiful dashboards with 40 charts that nobody opens after week two. The problem is never the tool. It's the design.
Design backward from decisions
Ask each user: what do you decide every week, and what number tells you? Build only those KPIs first.
- Plant head: OEE, tonnage, power per tonne
- Sales head: orders vs. target, pipeline, collections
- Purchase: stock days, pending POs, price variance
Rules I follow
- One page, one question
- The most important number goes top-left, biggest
- Every KPI shows a target and a trend, not just a value
- Colour only for exceptions: red means act now
If a chart doesn't change a decision, delete it.
Try the interactive dashboard on my homepage. Filter by year and segment, and every number recalculates.