A store generates a lot of data every day: revenue, receipts, average basket, margin, inventory, products or customers. A good store dashboard isn't about displaying as many figures as possible. It should let you quickly answer two questions: what is happening in my business? What deserves my attention? Choosing the right KPIs is therefore more important than how many there are.
Organize KPIs by goal
A dashboard quickly becomes hard to read when all the figures are presented at the same level. Organizing by goal lets you distinguish five main dimensions:
- sales;
- profitability;
- inventory;
- customers;
- day-to-day operations.
Each group answers a different question.
Sales KPIs
Revenue
It's naturally one of the first KPIs tracked. But displaying:
Revenue: €18,400
doesn't tell you whether the result is good or bad. It needs context:
Revenue: €18,400 — +6% vs. the previous week
Comparison turns a figure into information.
Number of receipts
The number of transactions helps explain how revenue changes. Two weeks can generate exactly the same revenue in very different situations. The number of receipts may have gone up while the average basket went down, or the other way around.
Average basket
The average basket is calculated like this:
Average basket = revenue ÷ number of receipts
It shows the average amount spent per transaction. Its trend is especially useful when analyzed alongside the number of receipts: it lets you tell a traffic effect from a change in the average value of purchases — see increasing the average basket.
Sales by family and by item
Overall revenue doesn't show which products explain its trend. Tracking sales by category or item lets you identify:
- the products that are growing;
- those that are slowing down;
- the categories that contribute the most to revenue;
- the items whose behavior changes sharply.
This is often where the explanation for an overall change lies.
Profitability KPIs
Revenue measures sales, not their profitability. Two periods showing the same revenue can produce different economic results depending on the products sold and the discounts granted.
Margin
When this data is available, margin completes the reading of revenue. A promotion can, for example, sharply boost sales while reducing the margin generated.
Product mix
The split of sales between products or categories can also change profitability. A rise in low-margin items and a drop in more profitable ones can change the economic result without any dramatic change in revenue. The dashboard should therefore show what is sold, not only how much is sold.
Inventory KPIs
An inventory dashboard isn't meant to display the thousands of quantities available in the catalog. It should mainly bring out the situations that need checking: see Inventory turnover and Stockouts.
Stockout risks
An item whose stock becomes insufficient relative to its sales pace deserves special attention. The remaining stock alone isn't enough: five units can represent several weeks for a slow product or a few hours for a highly demanded item.
Inventory turnover
Turnover measures how fast stock renews. It notably helps tell the items that move quickly from those that tie up stock for a long time.
Dead stock or overstock
A product can be available in large quantities while selling very little. Bringing these items out keeps problematic stock from staying invisible simply because it triggers no stockout — see dead stock. The inventory dashboard should therefore watch both extremes: not enough stock and too much stock.
Customer KPIs
When sales are linked to identified customers, several KPIs become useful.
Purchase frequency
It tells you how often customers come back.
Recency
The date of the last purchase helps spot customers who are starting to drift away.
Amount spent
It lets you identify your best customers, those who contribute the most to revenue over a period.
This data is best looked at together rather than separately. A very important customer isn't necessarily the one with the biggest receipt: regularity and frequency count too.
Compare periods
A dashboard should almost never show a KPI without a reference. Depending on the need, you can compare:
- today to yesterday;
- this week to the previous one;
- this month to the previous one;
- a period to the same period last year;
- the actual result to a target.
The choice depends on the business. Comparing a Saturday with the previous Friday isn't necessarily relevant if the two days behave very differently commercially. Likewise, a promotional week shouldn't be used carelessly as the reference for a normal week. A good dashboard therefore doesn't only show a gap: it lets you put it in the right context.
Compare several locations
In a network, the same KPIs can be compared across stores: see Multi-store management. But the goal isn't simply to build a ranking. A location may have lower revenue because it is smaller, while posting a particularly strong average basket or inventory turnover. Gaps mostly serve to raise questions:
- Why does this family perform better in this store?
- Why does this item have more stockouts here?
- Why is the average basket changing differently between two locations?
Comparison then becomes an analysis tool rather than a league table.
Operational alerts
Not everything useful on a dashboard is a KPI. Some information mainly tells you what needs to be checked or handled. Depending on the business and the data available, this could be, for example:
Minimum stock reached
An item is approaching its reorder threshold.
Use-by date approaching
A product with a use-by date must be monitored before it becomes unsellable.
Overdue invoice
A missed due date may call for a check or a reminder.
Quote pending
A quote sent a while ago without a reply may need a follow-up.
Cycle count due
Part of the stock needs to be checked as part of a scheduled cycle count.
These items don't all have the same importance or urgency. The whole point of the dashboard is to bring out the relevant exceptions rather than forcing the merchant to check every tool separately.
How many KPIs should you display?
There is no universal number. A dashboard with 30 KPIs isn't necessarily better than one with 8. Each KPI should answer a useful question. If nobody knows what decision to make when a figure changes, its place on the dashboard deserves to be questioned. A simple approach is to ask, for each piece of information:
- Why am I looking at it?
- How often can it change?
- What will I do if it changes sharply?
This helps you avoid dashboards filled with data that is never really used.
A dashboard shouldn't show everything at the same pace
Not every KPI needs to be checked daily. Some are useful every day:
- revenue;
- receipts;
- urgent incidents or alerts;
- major stockouts.
Others lend themselves better to a weekly look:
- average basket;
- family trends;
- turnover;
- customers who are starting to lapse.
And some KPIs can be analyzed over longer periods. The right pace mostly depends on how fast a piece of information can lead to a useful decision.
From dashboard to analysis
A dashboard lets you spot a signal. But it doesn't necessarily explain the cause. Imagine:
Revenue: -8%
The dashboard shows there is a gap. The next step is to ask:
“Why did my revenue drop this week?”
The analysis can then look at the number of receipts, the average basket, the families concerned or the items that contribute most to the change. You thus move from: see → understand → decide. This is an important distinction between tracking KPIs and analyzing sales.
How does AI complement a dashboard?
AI doesn't necessarily replace the dashboard. The two serve complementary uses. The dashboard is effective for quickly seeing the KPIs you track regularly. AI becomes especially useful when you need to dig into a question:
“Summarize the key points of this week for me.”
Then:
“Why did my revenue drop?”
And next:
“Which items mainly explain this drop?”
The analysis can thus evolve question by question without having to plan every filter or report in advance.
How can Gillia personalize this tracking?
With Gillia, a request that is useful day to day can be pinned. It can then run automatically to feed a dashboard tailored to the store's priorities: this is the principle of routines. For example:
“Every morning, show me yesterday's revenue, my average basket and the products close to a stockout.”
Another store may prefer to track:
“My sales by location, the products with no sales in 30 days and the customers to follow up with.”
The point is not to impose the same dashboard on every store: the information tracked can match the user's real priorities. Gillia can also be asked when a KPI deserves deeper analysis.
From signal to action
A KPI is only of interest if it can lead to a decision. An item close to a stockout can lead to preparing a restocking. A product whose sales are slowing durably may need a look at its stock, at the risk of becoming dead stock. A customer who no longer comes back can become a follow-up target. A drop in revenue can lead you to look for the families or items responsible for the gap.
The useful sequence is therefore: KPI → detection → analysis → decision → action. The action shouldn't be automatic in every case. When it actually changes the business — an order, a promotion, customer communication — it should remain controllable and subject to approval.
To go further, explore Gillia's uses: reading dashboard alerts or triggering a quick action from an alert.
To go further, discover the Gillia POS: every sale updates inventory, customers and the online store.