In short: If you cannot answer which customers are growing, the reason is usually that the data was never collected, not that anyone was careless. The fix is not a CRM project. It is one spreadsheet, three years of invoice lines and a definition agreed before anyone opens it.
Last week I wrote about how to define your target market and listed the questions your own finance data can answer. Since then, several owners have said a version of the same thing: fine, but I cannot answer a single one of them. We do not have the data.
That is the honest answer. It is also the most common one.
The data was never collected because nobody needed it
The invoice system records what was sold, to whom, at what price. It was built to close the month and report VAT, not to show whether a customer group is structurally growing.
In a company of twenty, forty or sixty people, the commercial memory is a person. Usually the owner. That person knows which customers are quiet this year, which deal nearly went to a competitor and which product line is quietly taking off. Memory is fast, free and it works, until the day you need it in a board meeting, a strategy process or a due diligence room.
Nobody decided not to collect data. There simply was no need for it. Until there was.
What "we have no data" usually means
In practice it is rarely that nothing exists. It is that what exists cannot be trusted for a decision:
- The same customer appears under three spellings: the legal name, the trading name and whatever the person in accounting typed that week.
- The customer field sometimes holds a project name rather than a company.
- A product group called "other" carries a quarter of the revenue.
- Quotes live in someone's inbox. Lost deals exist nowhere.
- The reference list is an Excel file that doubles as a marketing brochure: the largest logos, no dates, no values, no repeat pattern.
- Nobody can say when a given customer first bought, or how often they come back.
None of this is a data-quality scandal. It is the normal condition of a company that grew through relationships instead of through systems.
Why it costs money
Without a baseline you cannot tell a trend from a good year. Energy was busy in 2025. That does not mean your energy segment is growing structurally; it may mean two large projects landed in the same twelve months.
Without a baseline you cannot see repeat business, which is the most reliable signal of where you are genuinely good. And you cannot see that the growth of the last two years came mostly from two customers, which is not growth but concentration.
A market definition made on top of that is a guess with a nicer name. Which is how boards end up approving a market that is really one attractive project, or twelve areas that came out of a meeting and now have to be reflected in the website, the sales material and the hiring plan at the same time.
The fix is smaller than it sounds
It is not a CRM implementation. It is not a data warehouse. It is a baseline, and it can be done in two to four weeks, part-time, by somebody who knows the business.
- Agree the definitions first. What counts as one customer: the invoice address, the group, the contact? What counts as a sale: net invoiced or booked orders? What counts as repeat: a second invoice within how many months? Do this before opening the spreadsheet, or the result will be argued about afterwards.
- Export three full years of invoice lines, plus year to date. Lines, not reports: date, customer, item or product group, quantity, unit price, net amount, credit notes.
- Merge the customer names into one column. This is the step where the insight appears. Most surprises are not about the numbers; they are about discovering which names turn out to be the same company.
- Add five columns per customer: first invoice date, last invoice date, number of invoices, revenue last year, revenue this year.
- Sort by revenue and look at the top twenty. For each: who else in the company has sold to them, and is there a second company like them we have never approached?
- Add the quote ledger, three years with outcomes. Deals you lost and deals you never quoted for are the part finance cannot show you.
- Write down what this still cannot answer: who decided, why they chose you, what would have made them buy more. That list becomes your conversation list.
Be honest about what you now have
A baseline tells you what happened. It does not tell you why. It cannot reveal that a customer stayed because of one engineer, or that a promising segment is real but buys on a three-year cycle. Finance narrows the field. It does not name the market on its own.
And if nobody in the leadership team can produce this baseline within a month, that is itself a finding. It means the company is running on memory, which is the same exposure as the founder being the company, only spread across a few heads instead of one. It is also a good argument for having someone at board level who knows which question to ask the data.
If you would like the one-page checklist we use before we start mapping your market, write to me and I will send it over.