Picture an owner, call her Marta, who runs two bakeries in a mid-sized town. She is at the first one by half past five, and by seven she has looked at exactly three numbers. Not revenue, not profit, not the bank balance. Those she checks on Monday mornings with a coffee and a calculator. The daily three are different, and the reason she chose them is the most useful lesson in small-business measurement we know.
Her numbers are: yesterday's waste as a share of what was baked, the number of pre-orders for the next two days, and how many of yesterday's customers paid with a loyalty card.
None of them is money. All of them tell her something about money before it arrives.
Lagging numbers are a report card
Revenue is the number every owner watches, and it is the least useful one to watch daily. By the time revenue drops, the reasons for the drop are weeks old. A new competitor opened, a regular stopped coming, the croissants got slightly worse when a new baker started. Revenue tells you that something happened. It rarely tells you what, and never in time.
Accountants call these lagging indicators: they record the result of decisions already made. They are essential for the monthly review and useless for the morning.
Leading indicators are the opposite. They move first. They are noisier and less precise, but they give you days or weeks to react. The trick is that they are specific to your business. There is no universal list, which is why most dashboards default to the lagging numbers everyone understands.
Why these three
Waste. In a bakery, unsold bread is the single largest controllable cost. Marta's target is under 6% of baked goods. When it creeps to 9% for three days, it means she is baking for customers who are no longer coming, usually on a particular weekday. She adjusts the next day's plan, not next month's.
Pre-orders. Cakes and catering orders arrive two to ten days ahead. A thin order book on Tuesday means a weak weekend. That gives her time to post something on the local Facebook group, call the two offices that order sandwich platters, or simply schedule one fewer person on Saturday.
Loyalty-card share. This is her early warning for the thing that actually kills small food businesses: regulars drifting away. If 55% of transactions usually come from card holders and it falls to 45% over a fortnight while total customers hold steady, she is replacing regulars with passers-by. Passers-by are fine, but they do not come back in February.
Finding your own three
You do not need a consultant to find leading indicators. You need to work backwards from the last bad month.
Take the worst month of the past year and ask what you would have needed to know six weeks earlier to soften it. A design agency might answer: the number of proposals sent. A plumbing firm: the share of calls that turn into booked jobs. An online shop: returning visitors as a share of all visitors, or the conversion rate of people who added something to the basket.
Then check whether you can measure it without heroics. A leading indicator you cannot read every day or week is just a good intention. Marta's loyalty share comes straight out of the till system. Her pre-orders are in a paper book, which she counts. Her waste is weighed by the closing staff, written on a sheet and photographed.
Finally, give each number a range, not a single target. Waste between 4% and 6% is normal. Above 8% is a signal. Below 3% is also a signal, because it usually means the shelves were empty by four o'clock and she lost sales she never saw.
What the three numbers are not
They are not a strategy. Marta's strategy is to be the place people buy their weekend bread, which is why loyalty matters more to her than footfall. The numbers come from that choice. A bakery next to a train station would pick different ones: queue time at peak, perhaps, and sell-through of grab-and-go items before nine.
This is the part most measurement advice skips. A KPI only means something relative to a goal, and a goal only means something relative to a decision about what kind of business you are running. If you start from the dashboard, you end up measuring what is easy. If you start from the decision, you measure what matters and find a way to make it easy.
The weekly layer
On Monday, the three morning numbers get a summary line each, and the lagging numbers join them: revenue against the same week last year, gross margin, labour cost as a share of revenue. This is where Marta compares the two shops and decides whether anything structural needs to change.
The monthly layer is for the questions that need more than a week of data: price changes, whether the new sourdough is pulling its weight, whether to open on Mondays in summer.
Three layers, a handful of numbers in each, and a clear rule for which decisions belong to which layer. That is the whole system. It fits on one sheet of paper, and for years that is exactly where it lived.
The paper stops working at about the moment a second location opens, because the owner can no longer stand at both tills. That is the point where it is worth moving the same three layers into a tool. In ExploreKPI each number gets a target and a minimum, the morning ones are entered daily and roll up into the weekly view on their own, and both shops sit side by side. The system does not change. The morning read just takes ninety seconds instead of a walk across town.
