TL;DR: Eight metrics on one page, refreshed weekly, with sparkline plus variance to plan plus variance to prior period. That is the dashboard people actually read. Everything else is decoration. If your dashboard has 30 metrics, it has zero readers.
The reason most finance dashboards fail is not the tool. It is the count. Thirty metrics on a Looker page, or four tabs in a Tableau workbook, is not a dashboard. It is a data dump. Nobody outside of FP&A opens it twice.
Executives read dashboards under time pressure, usually on a phone, usually between meetings. The dashboard has to answer three questions in under 45 seconds: is the business on track, what is off, what should I ask about. If it does not answer those three questions on the first screen, it fails.
The 8 metrics that belong on a one-page dashboard
Every business is different. But across most operating companies, the same eight metrics do the work. Pick these first, add a specialized ninth or tenth only if the business truly needs it.
| # | Metric | Definition | Why it earns its slot |
|---|---|---|---|
| 1 | Revenue growth (YoY) | Trailing 3-month revenue vs same period prior year | The one number the board opens with, every time |
| 2 | Gross margin | Revenue minus COGS, as a percent, trailing 3 months | Tells you if the business model is holding under growth or pricing pressure |
| 3 | Cash on hand | Operating cash plus available revolver headroom | The number that determines how many options you have next quarter |
| 4 | DSO minus DPO | Days sales outstanding less days payables outstanding | Working capital drag in one number. Moves the credit line more than P&L noise |
| 5 | Headcount cost as percent of revenue | Total loaded people cost divided by trailing 3-month revenue | Fastest single indicator of operating scale kicking in, or breaking down |
| 6 | Pipeline coverage | Qualified pipeline for the next 90 days divided by the same 90-day sales target | Forward-looking, not lagging. 3x coverage is healthy for most B2B, more for consumer |
| 7 | Top customer concentration | Top 1 customer share of revenue, and top 5 share of revenue | The risk number the board asks about after every good quarter |
| 8 | One business-specific leading indicator | Varies. Bookings for SaaS. Same-store sales for retail. Utilization for services | The one metric that predicts the next 60 to 90 days for your specific business |
Each metric shows a 12-month sparkline, the current value, variance to plan, and variance to prior period. That is four pieces of information per metric, on one line. Eight lines. Fits on one screen.
Visual template description
Imagine a single page divided into two columns and four rows. Each cell holds one metric. Inside a cell:
- Metric name at the top, bold, 14pt.
- Current value centered, large, 28pt.
- Variance to plan and variance to prior period on either side of the current value, colored (green if favorable, red if not), 12pt.
- 12-month sparkline under the number, thin line, no axis, no legend.
No pie charts. No 3D bars. No gradient fills. If a colleague cannot read the page from three feet away, redesign.
The 22 metrics you can stop tracking
These live on the “reports we send but nobody reads” list. They are not wrong. They are just not weekly. Track them monthly in a longer pack, or on demand.
- Revenue by product line, weekly
- Revenue by region, weekly
- Bookings by rep, weekly
- Marketing qualified leads count
- Cost per lead by channel
- Website conversion rate
- NPS, weekly
- CSAT, weekly
- Time to close (support tickets)
- Deal velocity by stage
- Average selling price by rep
- Discount rate by segment
- Trial to paid conversion, weekly
- Backlog by quarter of expected recognition
- Aging AR buckets over 30 days, weekly
- Vendor spend by category, weekly
- Real estate cost per square foot
- IT spend per FTE
- Software spend per FTE
- Recruiting cost per hire
- Employee turnover, weekly
- Training hours per FTE
Every one of these has a use. None of them belong on a weekly one-page dashboard read by the CEO and board.
Cadence: who sees what, when
- Weekly, Monday 8am: The one-page dashboard goes to CEO and functional leaders. 8 metrics, no attachments.
- Monthly, day 8: Full monthly reporting pack. 20 to 30 metrics organized by function. Read by CFO, CEO, VP of Finance, function leads.
- Quarterly, week 3 of new quarter: Board pack. Includes trailing 4-quarter view of the same 8, plus deeper narrative on any that broke plan.
Notice the one-pager is read weekly. That is what earns the sparkline and the design attention. The monthly pack is read once. The board pack is read once and referenced once. Design accordingly.
How the 8-metric dashboard fails
Three ways this dashboard breaks in year 1:
- Definition drift. Nobody agrees what “pipeline” means. Sales says one thing, marketing says another, finance says a third. Write the definitions on the page footer, in 8pt. Do not skip this. Definition drift kills more dashboards than data issues.
- Data lag. If the dashboard runs on Monday but the data is only current through Thursday, executives quietly stop trusting it. Aim for data current through Sunday night by Monday 8am. If you cannot, publish Wednesday instead of Monday.
- Green everywhere. If every variance is green, either the plan was sandbagged or the thresholds are too generous. Recalibrate. A useful dashboard shows red 20 to 30 percent of the time.
The one business-specific metric
Slot 8 is the flexible one. It is the metric that predicts your specific business’s next quarter. Some examples that earn the slot:
- B2B SaaS: Net new logos in the trailing 30 days, or expansion ARR trailing 30 days.
- E-commerce: Repeat purchase rate for the trailing 90 days.
- Restaurant or hospitality: Same-store sales, week over week and vs prior year.
- Professional services: Billable utilization trailing 30 days.
- Manufacturing: On-time in-full delivery rate, or backlog in weeks of production.
- Marketplace: Take rate or GMV per active user.
Pick one. Not two. If you have a strong argument for a second, that means one of the other seven slots is not earning its place and should be replaced.
What to hand executives with the dashboard
Nothing. That is the point. The dashboard is the deliverable. If it needs a 3-paragraph email to be understood, it is not a dashboard.
If a specific metric goes red, add one line at the top of the page. “GM off 180bps this month, driven by mix shift into lower-margin segment. Deep dive attached.” One line. Not a memo.
How to pare a bloated dashboard down to 8
Most CFOs inherit a dashboard with 25 to 40 metrics on it. The right move is not to argue for cuts in a working session. It is to run a two-week silent test.
- Week 1: Instrument. Add a simple usage tracker to the dashboard, or if the tool does not support it, ask three executive readers to note which metrics they actually looked at that week.
- Week 2: Compare notes. Almost always, executives report reading 6 to 10 metrics. The others live on the page but are never opened.
- Week 3: Publish the pared version. Do not ask permission to cut. Publish the 8-metric page in parallel with the old dashboard. After four weeks, the old dashboard will feel bloated and readers will migrate on their own.
Do not announce the cut. Let usage do the talking. The old dashboard can retire quietly after a month.
What the 8-metric page does not do
It does not replace deep-dive analysis. It does not replace the monthly reporting pack. It does not replace the board deck. It is a health check. It exists so executives can spend 45 seconds a week on the whole business and know where to spend the next 15 minutes of attention.
Everything else in a good reporting stack still has a place. The one-pager is the entry point, not the total system.
Push back on this.
Every operator’s situation is a little different. If you run this differently, disagree with the methodology, or think we got something wrong, tell us. We publish the best counter-approaches on our Reader Contributions page, credited or anonymous, your call. Email hello@thepragmaticcfo.com.
FAQ
What tools should I use to build this?
Whatever your team already knows and can maintain. Google Sheets with a scheduled data refresh works. Looker Studio works. A native dashboard in your ERP works. Do not switch tools to build the 8-metric dashboard. Build it in what you have.
How do I get the CEO to actually read it?
Send it at the same time every week (Monday 8am is best). Make it look identical every week. Change nothing about the format for at least a quarter. Executives read what is predictable. Novelty kills readership.
Should I include forecasts on the dashboard?
Not on the weekly page. Forecasts belong in the monthly and quarterly packs. The weekly one-pager is a health check. Health check plus forecast plus commentary is a monthly deliverable.
What about dashboards for functional leaders?
Different page, different metrics. Sales should read a sales dashboard, ops should read an ops dashboard. The 8-metric page is for the CEO, CFO, and board. Do not merge audiences on one page. Ever.
Do I need real-time data?
Almost never. Cash and pipeline are worth intraday freshness if there is a live capital or sales event. Everything else can be as of end of prior day, or even end of prior week. Real-time dashboards are almost always a vanity project.
Related reading
- The PE-Backed CFO Board Reporting Package
- The FP&A Budget Cycle That Actually Works
- Rolling Forecasts vs Annual Budgets
- The AI-Assisted CFO Weekly Cadence
- The 5-Prompt Weekly Financial Review
Sources
- AFP FP&A Benchmarking. afponline.org
- AICPA CGMA finance function research. aicpa-cima.com
- McKinsey research on management reporting. mckinsey.com
Written by The Pragmatic CFO. 15+ years running FP&A and building AI-native finance workflows across portfolio companies.
The Pragmatic CFO 