Your Sales Dashboard Needs a Conversation

A number is a signal, not an explanation

Imagine a family venue normally has a busy café after its morning sessions. This Tuesday, the dashboard shows noticeably lower café sales. It is tempting to jump straight to a conclusion: the team did not sell enough, the offer is wrong, or Tuesday staffing needs to change.

Now imagine what the report does not show. One session finished early because of a timetable change. A group booking included refreshments that had been paid for previously. The café also ran out of one frequently ordered item during lunch. Each could influence the sales figure, but each points towards a different operational question.

This is why a dashboard works best as the start of an investigation rather than the end of one. Use unusual figures to identify where you should look, then gather enough context to understand what customers and staff were actually doing.

Resist the instant story

People are good at explaining a graph quickly. Sales rose after a menu change, so the menu change must have worked. Sales fell during a particular shift, so the shift must have performed badly. Those explanations may be true, but the graph alone has not established them.

Before acting, write the observation without the explanation: “Sales between 2pm and 4pm were lower than in the comparable period.” Then investigate why. Keeping observation and interpretation separate makes it easier to notice evidence that challenges your first assumption.

Check what the report is actually measuring

Before comparing two figures, make sure you understand what each figure represents. Is the report based on orders entered, completed sales, payments received or another definition? How are refunds, deposits or later payments represented? Use the documentation or support information for your own reporting setup rather than assuming every report treats them identically.

Check the periods too. Comparing Monday to Monday sounds reasonable until you discover that one Monday included eight trading hours and the other included six. A bank holiday, private event, delayed opening or early close can make superficially similar dates quite different.

Bookings can complicate the picture further. A customer may reserve an event on one date, pay a deposit on another and visit weeks later. Depending on what you are reviewing, the date of customer activity and the date money appears in a particular report may not tell the same story.

Ask staff about a specific time and place

“Why were sales bad on Tuesday?” is a poor operational question. It is broad, loaded and invites people to defend the day rather than describe it. Ask something staff can answer from what they observed.

Try: “What was different between 2pm and 4pm on Tuesday compared with a normal Tuesday?” A colleague may remember that the first afternoon session was quieter, customers queued at reception during the usual café period, or a commonly requested product became unavailable.

Then make the question more precise. If someone says “it was quiet”, ask whether they mean fewer people entered the venue, fewer customers approached the counter or the same number of people bought less. Those are different situations.

Keep the conversation about the operation, not blame

If staff believe every dashboard discussion is a performance accusation, useful context may disappear. Frame the review around understanding what happened and deciding whether anything should change.

Ask colleagues what customers requested, where queues formed, what was unavailable, what unusual bookings occurred and whether normal opening or session times changed. Record facts that help interpret the report. Avoid turning unsupported impressions into established explanations.

Put a second measure beside the first

Total sales tell you how much was recorded under the definition of the report you are using. They do not tell you whether the change came from customer numbers, purchasing behaviour or something else. A second relevant measure can narrow the question.

If available and appropriate in your reporting setup, compare transaction count with total sales. Fewer transactions with a broadly similar amount per transaction points towards a different investigation from a similar number of transactions with a lower average value.

But keep the limitations in view. A transaction is not necessarily a customer. One family might pay together, while another makes separate purchases for admission, food and drinks. Use transaction counts as one clue rather than casually relabelling them as visitor numbers.

Use booking information when the business is booking-led

For venues built around sessions, classes or reservations, look at the relevant booking activity alongside sales where appropriate. A quieter café period may make more sense when you discover the preceding session had fewer attendees. Conversely, strong attendance with weak secondary sales suggests a different question about what happened during the visit.

Keep the comparison logically connected. Do not add more metrics simply because they are available. Choose the next measure because it helps distinguish between plausible explanations.

Connect sales with the customer journey

Sales figures become more useful when you place them alongside the journey customers took through the venue. Ask where customers were during the period and what opportunities they actually had to buy.

Consider a hypothetical leisure venue where customers usually buy refreshments after a one-hour session. Sales from 11am to noon fall. Before changing café staffing, check the timetable. If the relevant session was moved and families now leave the activity area at 12:15, the purchasing opportunity may simply have moved to a different reporting period.

Or imagine a café within a venue records plenty of transactions but lower spending during one afternoon. Check what was available and what customers were asking for. If a higher-priced popular item was unavailable, product mix may help explain the change. If the full offer was available, look elsewhere rather than forcing that explanation.

Walk the period, not just the spreadsheet

For an important recurring dip, spend time in the venue during the relevant period. Observe when customers arrive, where they wait, whether they can easily see the offer and whether queues or session timings affect their opportunity to purchase. Compare those observations with the figures afterwards.

This does not replace reporting. It gives the report operational context. A line on a chart becomes much more actionable when you can connect it to a specific part of the customer journey.

Turn an explanation into a testable decision

Once you have a plausible explanation supported by the available evidence, choose an action that addresses it directly. Avoid changing prices, staffing, opening hours and the offer simultaneously unless there is a separate reason to do so. Too many changes make the next report difficult to interpret.

Suppose staff records show that the same popular café item was unavailable during several comparable busy periods. The practical action might be to review how that item is prepared or replenished before those periods. Then compare similar trading periods after the change and check whether availability improved as intended before making broader claims about sales.

If the issue was reduced opening time, comparing whole-day totals alone may continue to mislead. Add a comparison that accounts for the actual hours traded, while still noting that different hours can attract different patterns of customer demand.

Write down what would change your mind

Before testing an operational change, record what you expect to observe if your explanation is right. For example: “We think the queue between sessions is reducing café purchases, so we will adjust the counter workflow and observe the next four comparable session changeovers.”

Afterwards, review both the operational observation and the relevant reporting. If the queue improves but sales do not change, that does not make the operational improvement worthless. It does mean you should not claim the queue was the sole explanation for the sales pattern.

Build context into the sales review routine

You do not need a lengthy investigation every time a number moves. Create a short review habit that catches obvious differences before the team starts inventing explanations.

A practical sales-review checklist

1. Define the number: confirm what the report measures and how the period is defined.
2. Check comparability: compare trading hours, day type, sessions or other relevant operating conditions.
3. Look for obvious exceptions: note large bookings, unusual payments, closures, unavailable products or other known events.
4. Add one useful measure: choose transaction, booking or other relevant information that helps test the possible explanation.
5. Ask the people who were there: use a specific question about the time and part of the operation you are reviewing.
6. Choose an action only when needed: make the action address the evidence rather than the graph in general.
7. Review it later: compare a suitable period after the change and record what you learned.

Over time, retain brief context for genuinely unusual periods. “Private function; normal public service ended at 6pm” will be far more useful months later than trying to remember why that evening looks strange on a chart.

Dashboards are valuable precisely because they make patterns visible. The mistake is expecting the reporting layer to contain every explanation for human behaviour, customer flow and day-to-day operations.