The trouble with most dashboards
Many business owners have access to a dashboard: the one in their accounting software, the one in their POS, one somebody on the team put together. And many of them rarely open it — not because they don't care about their numbers, but because the dashboard doesn't answer what they need to know.
The symptom is familiar: a screen full of charts, big numbers with no context, colors whose meaning nobody remembers. You glance at it, half understand it and go back to asking for the Excel report.
The fix isn't more charts. It's organizing information around the questions an owner or manager actually asks. After building dashboards for very different businesses, we use four.
The four questions
- How are we doing?
- What's happening?
- Where's the problem?
- What do I need to decide?
Each question maps to a band on the dashboard, top to bottom. The first takes seconds to answer; the last tells you what to do. Let's go through each.
1. How are we doing?
The first band shows a few KPIs — three to five — that summarize the health of the business for the period. For example: revenue, gross margin, operating income and cash on hand.
The most important rule for this band is that no number stands alone. Monthly revenue, by itself, says nothing. It says something when it's compared: with last month, with the same month last year or with the goal you set. If a KPI has no comparison, you can't tell whether it's good or bad.
What to look for
- Are the headline KPIs above or below target?
- Is the gap large or small?
- Is the month complete, or is data missing? A good dashboard tells you when a period is preliminary.
If everything is within range, you can stop here. If something stands out, move down to the next band.
2. What's happening?
The second band shows the trend. One number can be a fluke; a trend is a signal. Here you see the same KPIs from the first band across several months or weeks.
You're looking for direction and timing. Has margin been sliding for three months, or did it drop all at once this month? Is revenue growing while expenses grow faster? That distinction changes the whole conversation.
A caution: trends get distorted when the data isn't comparable. If one month includes a large equipment purchase booked as an expense, operating income will look worse than it really was. A well-built dashboard keeps capital spending and inventory purchases separate from operating expenses.
3. Where's the problem?
The third band breaks things down. If margin fell, which location, which service line, which channel, which client? This is where information becomes actionable.
At a restaurant, overall margin may have dropped because delivery sales grew, and delivery carries commissions. At a clinic, one procedure type may have a higher cost per case than quoted. At a distributor, a large customer may be paying later and later.
What to look for
- The segment that best explains the change you saw above.
- Big gaps between similar segments (two locations, two providers, two channels).
- Segments growing in sales but not in margin.
If you operate in more than one currency, make sure the dashboard doesn't add them together directly. Dollars and pesos should be shown separately or converted at a dated, visible rate.
4. What do I need to decide?
The last band is the one most dashboards are missing. It shows what needs your action: alerts, open items and variances someone has to handle.
Examples: clients past a certain number of days overdue, locations under target, expenses over budget, inventory below minimum, months not yet closed. Every alert should come with the data that triggered it so you can verify it.
This band turns the dashboard into a management tool. Instead of staring at numbers and wondering what they mean, you get a short list of decisions.
Three signs you can trust your dashboard
The four questions only work if the numbers are right. Before making decisions, check for these three signs:
- Every KPI explains its formula. Hover over or tap a KPI and it should tell you, in plain language, how it's calculated.
- The dashboard tells you what's missing. If a month doesn't have payroll or the bank balance loaded, it should say so. Showing a 100% margin because costs are missing is worse than showing nothing.
- Totals tie to the detail. Locations should add up to the total. Dashboard revenue should match your books.
If your current dashboard fails any of these, the decisions you make with it may rest on incomplete numbers.
A ten-minute routine
With a dashboard organized this way, a weekly review can take a few minutes:
- Look at the first band. Anything unexpected?
- If so, move to the second: is it a one-off or a trend?
- Move to the third to find where it's happening.
- Check the fourth band and assign each open item to someone.
Once a month, at close, run the same review more carefully and adjust targets if something in your business has changed.
Common mistakes when reading a dashboard
Even with a good dashboard, some reading mistakes come up again and again. Knowing them helps you avoid rushed decisions.
Reacting to a single data point
One bad month isn't always a trend. Before changing prices, cutting costs or calling out a team, check the second band to see whether the change holds over time or was a one-off — a holiday, an unusual purchase or a client who paid late.
Confusing revenue with cash
Selling isn't collecting. A month with strong sales can come with very little cash if clients pay on credit. Look at the P&L and cash flow together, and don't draw conclusions about liquidity from sales alone.
Comparing periods that aren't comparable
Comparing December with January at a restaurant, or a month with two payroll runs against one with three, distorts the picture. Where seasonality matters, compare with the same month last year as well as the prior month.
Mixing owner draws with expenses
Owner draws aren't a business expense. If they're mixed in with operating expenses, results look worse than they are and cost-cutting decisions get made on the wrong basis.
Ignoring the incomplete-data warning
If the dashboard says a month is preliminary or a source is missing, any conclusion about that month is provisional. Wait for close, or load the missing data, before deciding.
Treating every alert as urgent
Not every flag in the fourth band needs action today. Sort alerts by impact — how much money or how many clients are involved — and deal with the biggest first. A dashboard that raises too many alerts needs its thresholds tuned, not more attention.
The bottom line
A dashboard isn't measured by how many charts it has but by how many decisions it helps you make. Organizing it around four questions — how are we doing, what's happening, where's the problem and what do I need to decide — is a simple way to get there.
Want to see what yours could look like? Our homepage has an interactive demo with fictional brands. And if you'd like one with your branding and your data, see our dashboards and KPIs service — or, if your challenge is your online presence, our websites service.