The method does not care what you sell. Take a defined set of transactions, follow every dollar to where it actually landed, and separate the money that was never collectible from the money that leaked. In a shop that means margin, inventory and card fees. In a service business it means job costs, recurring plans and receivables. Same arithmetic.
Real work, on real books
Not a mock-up. Every figure below can be checked against its published source.
Real analysis from a live engagement, published with the owner’s permission and de-identified: the business is not named, located or described in a way that identifies it, and it is independent of Belle Curve Insights and of its officers. Figures are unaudited.
A method sample rather than a client result: the retailer’s transaction log is published data, converted to US dollars at the 2011 annual average rate. Figures are unaudited.
This is the work I spent years doing for large brands — figuring out which customers drove profit and which quietly drained it, then deciding what to say to each group. The arithmetic does not change at smaller volumes. A business doing nine hundred transactions a month can answer these questions as well as one doing nine million, and can act on the answer the same week.
And when transactions don’t tie to a person
Plenty of businesses sell to people they never identify — cash sales, no loyalty scheme, no accounts. That rules out customer-level profiling, and anyone who tells you otherwise is guessing.
It does not rule out the analysis. It changes the unit from the customer to the basket, and for some businesses the basket is the more useful thing to study anyway:
| Basket-level question | What it changes |
|---|---|
| What sells together, reliably? | Bundles, placement, and what to put next to what |
| What does a given item pull along with it? | Which low-margin lines are worth keeping because of what rides with them |
| How does basket size shift by hour and by day? | Staffing, opening hours, and when to promote |
| What did that discount do to the rest of the basket? | Whether a promotion grew the order or just shrank the margin on it |
That last one is where a lot of money hides. A promotion is usually judged on whether the discounted item moved. The better question is what happened to everything else in the basket while it did — and that answer does not need a single customer name.
Which analysis is right depends on the business. A shop with steady anonymous footfall learns more from its baskets. A subscription service, a salon, or anything with repeat named customers learns more from the customer. Most businesses have some of both, and the analysis starts by working out which side your data can actually support.
How it works
One standard export from your accounting system is usually enough to start. No new software, no IT project, and nothing to install. The analysis is fixed-scope and fixed-fee, and you get a plain-English findings report with each item sized in dollars and ordered by what to do first.