How to build an executive KPI dashboard (that the leadership team actually uses)

An executive KPI dashboard is a single page that tells the leadership team, in a few minutes, where the business is on plan, where it is drifting and which gaps need a decision now. To build one the team actually uses, pick eight to twelve KPIs tied to this year's priorities, group them by perspective on one page, show actual, target, variance, trend and status for each, give every number a named owner and a written definition, and review it on a fixed rhythm. The software you choose matters far less than those five decisions.

Most companies with thirty to three hundred people are not short of data. What they lack is one page the leadership team trusts, reads the same way and acts on. That is a design problem, not a technology problem.

What an executive dashboard is for

A report answers the question “what happened?”. An executive dashboard answers a different one: “where do we need to act?”. If the leadership team can read the page and nothing changes in what they do that week, it is a report with nicer formatting.

A useful test is to take each KPI in turn and ask what the team would decide if it turned red. If nobody can answer, the number does not belong on the executive page. It may still be valuable on a team dashboard.

A dashboard that earns its place does three things:

• It shows whether the plan is on track, against the targets the leadership team committed to, not against last year or against gut feel.
• It surfaces exceptions early enough to act, which means monthly or weekly data, not a quarterly surprise.
• It creates one version of the numbers, so the meeting debates the response rather than whose spreadsheet is right.

How to choose 8 to 12 KPIs

Eight to twelve is a working range, not a rule. Fewer than eight and there are usually blind spots, often in operations or people. More than twelve and attention spreads so thin that nothing gets discussed properly. The practical limit is what a leadership team can review with care in the time it gives the dashboard each month.

Start from the priorities, not from the data you happen to have. The easy mistake is to open the reporting tool, see what is available and build from there. Instead, list the three to five things the company must achieve this year, the main risks to the plan, and the few numbers that tell you early whether each is going well.

A five-question filter

1. Is it linked to a company priority or a material risk this year?
2. Is there one person on the leadership team who answers for it?
3. Does it move within the review period, so that a change means something?
4. Can it be measured reliably, from a known source, without a week of manual work?
5. Would someone act differently if it moved?

A KPI that fails two or more of these questions goes back to team level. Aim for a mix of lagging indicators, which confirm results (revenue, margin, cash), and leading indicators, which predict them (pipeline, on-time delivery, open critical roles). If you are unsure where KPIs end and objectives begin, our article on OKRs vs KPIs draws the line, and the business KPIs every growing SME should track is a sound starting list.

It also helps to group KPIs by perspective. Robert Kaplan and David Norton made the case in their balanced scorecard work that financial measures alone give an incomplete, backward-looking picture, and proposed adding customer, internal process and learning perspectives. You do not need to adopt the full methodology to borrow the principle: a dashboard that only shows money tells you what has already happened.

The layout: one page, the same columns for every KPI

One page is a hard constraint. If the dashboard needs scrolling or a second tab, it has stopped being an executive dashboard. Group the KPIs under four or five perspectives (for most SMEs: financial, customers, operations and people) and put the same columns against every line.

Data ownership and the definitions sheet

Every KPI needs two owners, and they are often different people. The business owner sits on the leadership team and answers for the result. The data owner answers for the number being correct and on time. That way, the conversation about a weak month is about customers, not about CRM filters.

The tool that makes this work is a definitions sheet: one line per KPI, kept next to the dashboard and changed only by agreement. At a minimum it records:

• The name and a plain-language definition.
• The formula, including what is excluded.
• The source system and the cut-off date for each period.
• The business owner and the data owner.
• The target for each period and the status thresholds.
• The paired measure, if there is one.

The last line deserves attention. Goodhart's law, in its common paraphrase, says that when a measure becomes a target it stops being a good measure, because people optimise the number rather than the outcome. Pairing helps: put bookings next to margin, delivery speed next to rework, headcount growth next to leavers. The pair makes it harder to improve one number by quietly damaging another.

The review cadence

A dashboard without a fixed review rhythm decays within a quarter: numbers stop being updated and definitions drift. The fix is to tie it to the meetings the leadership team already holds.

Three habits keep it honest. The numbers close by the same working day every month. The dashboard goes out at least a day before the meeting, so everyone starts from the same facts. And no new numbers appear in the room. For how the weekly forum runs, see the weekly leadership meeting; for the quarterly one, our quarterly business review guide.

Excel or a BI tool?

Both work. The choice depends on where your data lives and who will maintain the dashboard, not on which looks better in a demo.

• A spreadsheet (Excel or Google Sheets) is fast to start, cheap and easy to change. The cost is manual updating and the risk of several versions circulating.
• A BI tool (Power BI, Looker Studio, Tableau and similar) refreshes automatically and lets owners drill into the detail. The cost is that it needs clean, connected sources and someone who owns the maintenance.

A sensible path for many companies is to run the dashboard in a spreadsheet for one or two quarters, until the KPIs and definitions stop changing, and then automate the ones that proved useful. Automating before the definitions settle simply makes the confusion faster.

Example: the Marés Altas dashboard

Marés Altas is a fictional 60-person company that sells, installs and maintains water-treatment equipment for hotels and industrial sites in Portugal and Spain. The figures are invented, purely to show the format.

Reading it takes a few minutes. Two lines are off track and they are connected: installations are slipping because two technician roles have been open for months, and the commercial director has slowed new contract signings to avoid promising dates the team cannot meet. The useful part of the meeting is not the reading. It is the decision that follows: approve a temporary subcontractor for installations, move the two technician hires to the top of the CEO's list, and review both lines again in four weeks.

Common mistakes

• Too many KPIs. Twenty-five numbers on a page means none of them gets discussed.
• Actuals with no targets. Without a target there is no variance, and without variance there is no status.
• Only lagging measures. Revenue and margin confirm what happened. They rarely warn you in time.
• Definitions that change quietly. If a formula changes, the history is restated and the team is told.
• A dashboard built by the data team alone. The leadership team has to choose the KPIs, or it will not own them.
• Reviews that end without actions. Every KPI off track should leave the meeting with an owner, an action and a date.

If you want a head start, our KPI dashboard template has the layout and the definitions sheet ready to adapt, and the company scorecard template covers the wider annual view. If the harder part is getting the leadership team to choose, own and review the numbers, that is the work of our KPI dashboards and performance management service. The first step is an Operating Assessment: we look at how you measure and review performance today, and what to change first.