Why your dashboards aren't decisions.
Walk into any enterprise and ask to see the data. You'll be shown dashboards — dozens of them, hundreds sometimes. Now ask a harder question: which decision did each one change last quarter? The room goes quiet.
Most organisations are rich in reports and poor in decisions. That isn't a tooling problem, and it isn't fixed by another chart. After seventeen years building data platforms for hospitals, insurers, ministries and manufacturers, the pattern is remarkably consistent: companies spend fortunes making data visible and almost nothing making it decidable.
The dashboard trap
A dashboard answers the question someone asked when it was built. Decisions are driven by the questions being asked now — in this board meeting, about this quarter, under this constraint. The gap between those two things is where analytics programmes go to die.
Worse, dashboards multiply. Every team commissions its own view of the truth, each calculated slightly differently, each defended in the room by the team that built it. The CFO's margin number disagrees with the GM's. Twenty minutes of every executive meeting is spent reconciling figures instead of acting on them. That isn't reporting — it's institutionalised doubt.
A number a leader won't act on without double-checking isn't insight. It's decoration.
What decision-grade actually means
Data is decision-grade when three things are simultaneously true:
- Trusted — there is one governed source of truth, with lineage anyone can audit. When the number appears, nobody asks whose version it is.
- Framed — it answers the question leadership is actually asking, in the shape they need to act: a recommendation, a threshold, a trade-off. Not a grid of fifty metrics.
- Timely — it arrives inside the decision window. A perfect answer three weeks after the board meeting is worth exactly nothing.
Miss any one of the three and the data reverts to wallpaper. This is why "more dashboards" never fixes the problem — it usually makes all three properties worse.
The method: Ground, Surface, Act
Getting to decision-grade is an engineering and governance problem before it is a visualisation problem. The sequence matters. Ground: engineer one governed, quality-assured source of truth — catalogued, lineage-tracked, trusted. Surface: turn that foundation into board-ready insight — the answers leaders ask for, not another chart library. Act: put AI agents to work on top, automating the analysis and reporting that used to consume headcount.
Organisations that skip Ground and jump straight to Surface get beautiful dashboards on untrusted data — the exact condition described above. Organisations that skip Surface and jump straight to AI get confident answers from a foundation nobody audited. The order isn't a preference; it's the difference between an asset and a liability.
The test
Here's the question to take into your next leadership meeting: what was the last decision this organisation made faster, or better, because of our data? If the answer takes more than ten seconds, your data isn't decision-grade yet — and no amount of additional reporting will change that. Changing the foundation will.