We've built management dashboards for clients ranging from a stone-crushing unit to a municipal development authority. Almost every first draft makes the same mistake: forty metrics on one page, and nobody reads past the first six. A dashboard that gets read isn't the one with the most data — it's the one where every number on screen one changes a decision.
Not a data problem — an editing problem. Each of these maps to one card on the dashboard below.
The one-question filter
Before a metric earns a place on the summary screen, run it through one test:
If the honest answer is no — it belongs in an appendix, not the summary.
This single filter removes most of the noise in a first draft. It's also why a dense dashboard is a warning sign, not a strength: it usually means no one has decided yet what actually matters.
The six numbers that usually survive
| Metric | Why it matters | Common mistake |
|---|---|---|
| Cash runway | Months of operating expense covered at current burn — not just the bank balance | Reporting the balance alone, with no burn-rate context |
| Revenue vs. plan | Shows whether the business is on track against its own budget | Comparing to last year instead — a different, often irrelevant, story |
| Overdue receivables | Ageing beyond agreed credit terms, as one number — the earliest sign of a collections problem | Dumping the full ledger instead of one summarised figure |
| Gross margin trend | Last 4-6 periods together, to separate a blip from a pattern | A single-period margin, which hides the trend entirely |
| 1-2 operational metrics | Business-specific — utilisation, occupancy, disbursement pace | Generic metrics copied from a template that don't fit the business |
| Direction flag | A simple up/down/flat marker on each of the above, at a glance | Making the reader calculate the trend themselves |
What the numbers should answer
Strip it back further and a board dashboard exists to answer just three questions in under a minute:
Everything else is detail a board member can ask for — it doesn't need to be on screen one.
What it looks like assembled
Here's the six-number framework above, built out as an actual one-page board dashboard — the kind we'd hand a client's board at month-end.
Gross Margin Trend — Last 6 Months
Receivables Ageing
Note the appendix link at the bottom of a real dashboard — full ledgers, ageing detail, and departmental splits live there, not on this page.
A layman's example
Where first drafts usually go wrong
The instinct to include everything comes from wanting to look thorough. In practice it does the opposite: it signals nobody has made the hard editing decisions yet. Cutting down to six or seven numbers is harder work than compiling every metric available — but it's the difference between a report that gets fifteen minutes of real attention and one that gets skimmed and set aside.
- Every metric on screen one passes the "would we act on this?" test
- Comparisons are against plan/budget, not just last year
- Trends are shown over 4-6 periods, not a single snapshot
- Detail ledgers live in an appendix, not the summary
- Operational metrics are specific to the business, not generic
Building it so it stays current
The best dashboard design is worthless if updating it takes a full day of manual pulls each month. Wherever possible, we build the underlying numbers to link directly to the accounting system or a live tracker, so the report becomes a byproduct of good bookkeeping — not a separate monthly project in itself.