What is customer count?

Customer count is the number of people who have actually bought from you: customers with at least one order that counts as a sale. It deliberately excludes registered accounts that never ordered and customers whose only orders were cancelled or refunded — this is a buyers metric, not an accounts metric.

Formula

customer count = COUNT(customers with ≥ 1 net-sale order)

Worked example

A store platform shows 9,400 customer accounts. The customer count here reads 7,150: about 2,000 accounts never completed an order, and a few hundred had every order cancelled or refunded. The 7,150 is the population every other customer metric describes — repeat rate, average LTV and the at-risk count are all shares or slices of exactly this set, so the numbers compose without caveats.

How Saldo Metrics computes it

The widget counts rows of v_customer_lifetime_value, which builds one row per customer with a non-null first order date from their net-sale orders (cancelled and refunded excluded throughout). The count is lifetime and unfiltered by period — a customer who last ordered two years ago still counts. Customers are as distinct as the store platform's customer records: the same person with two accounts, or two guest checkouts, counts twice, because there is nothing safe to merge them on.

Why it matters

Customer count is the base of unit economics: growth in orders can come from the same buyers ordering more or from new buyers arriving, and the difference decides whether acquisition or retention deserves the next euro. It is also the denominator that keeps repeat rate and average LTV honest — all three describe the same set.

Common mistakes

• Comparing against platform account counts. Registered-but-never-bought accounts inflate those; the gap is definition, not data loss.
• Reading it as active customers. The count is lifetime; pair it with the at-risk count and cohort economics for recency structure.
• Treating guest checkouts as unique people. Duplicate humans are counted per customer record; heavy guest-checkout stores overcount somewhat.

Where you see this in the app

The Total Customers KPI on the dashboard.