Perspective · 4 min read
Why dashboards go unread
Ask a mid-sized company how many dashboards it has, and the answer is often “quite a few”. Ask which ones are opened every week, and the list gets short. The unused ones were usually built with care. They simply were not built around a decision.
Start from the decision, not the data
The useful question is not “what data do we have?” but “what will someone decide differently after looking at this?” Which branch to visit this week. Which customers to call. Whether to approve overtime. A view that supports one such decision beats a page of twenty charts that support none.
Agree what each number should trigger
For every measure on the screen, write down who owns it, what “normal” looks like, and what happens when it moves outside that range. If nobody can say what action a number should prompt, it probably does not belong on the dashboard.
Trust comes before design
The fastest way to lose an audience is a figure that does not match the books. Before a dashboard is shared, its totals should be reconciled to a number finance already trusts. One mismatch in the first week can make a team stop looking for good.
Put it where people already are
Many dashboards fail because they require a login to a tool people never otherwise open. A scheduled email at 8am, a screen in the plant office, or a link in the weekly meeting agenda will often do more than a beautiful interactive report.
Retire what is not used
Usage can be measured. A view that has not been opened in a quarter should be questioned and, usually, switched off. Fewer, trusted views are worth more than a long list of forgotten ones.
This is the kind of work our business intelligence & visualization line does.