Most small businesses collect sales data. Very few turn that data into decisions. The gap between having data and using it well is almost always caused by the same handful of mistakes — mistakes that are easy to fix once you know what to look for.

These are the five most common errors, with practical corrections for each.

Mistake 1: Looking Only at Total Revenue

Total monthly or annual revenue is the number everyone checks first. It's also the least actionable metric in isolation.

Revenue of €50,000 in March can mean completely different things:

  • If March last year was €35,000 — strong growth
  • If March last year was €60,000 — significant decline
  • If margin on those sales was 8% instead of the usual 22% — a revenue number that hides a problem

The fix: always compare revenue against the same period last year, month-over-month, and against target. And segment it: revenue by product/service, by customer, by channel. A business where 80% of revenue comes from one customer is fragile even when the top-line number looks healthy.

Mistake 2: Confusing Correlation with Causation

Sales went up 30% in the month after you launched a new marketing campaign. Therefore the campaign caused the sales increase. Right?

Not necessarily. That same month, a competitor may have closed down. Or a seasonal pattern you hadn't noticed because you were looking at the wrong time period. Or a major client renewed a large contract that would have renewed regardless.

This mistake leads to doubling down on strategies that coincided with results but didn't cause them — and abandoning strategies that are working but whose impact isn't yet visible in the numbers.

The fix: before attributing causation, ask "what else changed during this period?" Use control groups when possible (run a campaign in one region, not all). Build a longer timeline — 12 to 24 months of monthly data starts to separate signal from noise.

Mistake 3: Analysing Data Without Segmenting It

Average metrics hide reality. An average order value of €250 tells you nothing if half your customers spend €50 and the other half spend €450. The strategies for those two groups are completely different.

This applies to every metric: average customer lifetime value, average sales cycle length, average conversion rate. The average smooths out the information that matters most.

The fix: segment before you average. Split customers by size, by product, by acquisition channel, by geography. The segments that emerge from this analysis often reveal where profit is actually concentrated — and where resources are being wasted on unprofitable segments.

Mistake 4: Using Data Only to Confirm What You Already Believe

Confirmation bias in data analysis: a business owner who believes Product A is the main driver of success will unconsciously filter analysis to support that belief — choosing time periods, comparisons, and metrics that make Product A look important, while ignoring evidence that Product B is growing faster and has higher margins.

This is the hardest mistake to catch because it's invisible when you're inside it.

The fix: define the questions before you look at the data. "I want to know which product has the highest margin per unit sold" is a neutral question. "I want to prove that Product A is the most profitable" is a biased question. The wording changes the analysis. Also: share your analysis with someone who will challenge your conclusions, not confirm them.

Mistake 5: Acting on Short Time Windows

A 15% drop in sales in one week triggers urgent meetings, strategy changes, and panic. Two months later, the data shows it was seasonal variation that happens every year at the same time — and the strategy changes caused more disruption than the "problem" would have.

Short time windows make noise look like signal. Decisions made on one week or one month of data are often decisions made on statistical noise.

The fix: determine the minimum time window for each type of decision. For weekly operational decisions, four to eight weeks of comparison. For strategic decisions, twelve to twenty-four months. Build a dashboard that always shows the same period last year alongside the current period. Seasonal patterns become immediately visible.

The Underlying Pattern

All five mistakes have a common root: analysing data reactively rather than systematically. When you open the spreadsheet only when something seems wrong, you're looking for an explanation for something that's already happened. When you have a structured dashboard with defined metrics and time comparisons, you see problems early — before they become crises.

The data analytics services at PC Data Insights include diagnostic review of existing data processes and dashboard development to replace reactive reporting with proactive analysis. See project examples in the portfolio or get in touch to discuss your situation.