Most enterprise dashboards fail in the first thirty minutes — not of building, but of thinking. Before any tile is placed, the BI team has either understood what the user is trying to decide, or it has not. This article makes the case for treating those first thirty minutes as the most important part of the project.
Why dashboards drift
Walk into any large insurer or bank and you will find dashboards that look professional, load fast, and do not answer the question they were built for. Numbers reconcile to the source. Filters work. The right colour scheme has been applied. The dashboard is well made. It is also unused, because the executive looking at it cannot find what they need to decide what they need to decide.
The cause is almost never tooling. Power BI, Tableau and Looker are all capable of building a useful executive view. The cause is that the BI team started building before they understood what the dashboard was for. Once a tile exists, momentum carries it through review. The dashboard ships. The dashboard does not get used. The team blames the executive for “not engaging”.
The thirty-minute rule
For every new dashboard, the BI team commits to thirty minutes of structured conversation with the eventual user before any visual is built. The conversation has three parts.
The first ten minutes — the decision. What is the executive going to do with this dashboard that they cannot do today? Not “see the data”. Not “monitor the portfolio”. Specifically: what is the action this dashboard enables or the question it answers? If you cannot finish the sentence “this dashboard exists so that ___ can do ___”, you are not ready to build it.
The second ten minutes — the cadence. When will this dashboard be looked at? Daily? Monthly? Only at quarter-end? The answer determines almost everything about its design. A daily ops dashboard has different requirements from a quarterly board summary. Build for the wrong cadence and the dashboard either drowns in detail or hides the signal.
The last ten minutes — the failure mode. What does it look like when the dashboard is wrong? What is the worst-case scenario the executive is worried about? “The number is wrong and I commit publicly to it” is a very different worry from “I miss a portfolio drift for three weeks”. Both are real. They lead to different designs.
What thirty minutes saves
We have done the maths on this for a few clients. A typical executive dashboard takes 4–8 weeks to build. About 30% of that time is rework after the first user-acceptance review reveals that the dashboard is not what the user wanted. That is two to three weeks of expensive BI engineer time, lost.
Thirty minutes of conversation, done before the build starts, removes most of that rework. The arithmetic is obvious. The reason it is not done universally is that BI teams are usually rewarded for shipping tiles, not for asking questions.
What it looks like in practice
The thirty-minute conversation is not a “discovery workshop”. It is one BI engineer in a room (or call) with the eventual user, with a notebook, asking three questions. It produces a single page of notes that the executive co-signs. The notes become the spec. If the dashboard does not match the notes, it is not finished.
For our clients, the format is even simpler than that. We use a one-page template — *decision · cadence · failure mode* — that the user fills in with us. It takes about half an hour. It saves weeks. Boring discipline, ordinary outcome.
The harder version
The thirty-minute rule is the easy version. The harder version, which we apply to executive packs and management information layers, is to refuse to build a dashboard until the decisions it supports are written down and signed off. Most of the time, that exercise reveals that the dashboard is not actually needed — the decision could be made from a small dataset that already exists, or it is being made well already, or nobody is going to act on the answer either way.
That is not a failure of the BI team. That is the BI team doing its job well. The most valuable BI work we have done at clients has been to retire dashboards, not to build them. The hardest sentence in BI is “you do not need this”.
If you are scoping a BI estate refresh, our Business Intelligence practice helps teams that want to ship fewer, better dashboards.