The number that reaches an investor should describe the business as it is on the day diligence begins, including cancellations. A larger customer count may buy a calmer meeting, but it creates a sharper problem when the underlying data appears two weeks later.
At 11:47 PM, Samira was at her kitchen table in Accra with a mug of tea gone cold beside her laptop. Her diligence call was scheduled for the morning. The investor had asked for a customer export, monthly revenue, churn, and a short note explaining how she defined an active account.
The dashboard said 184 customers.
Then she removed the accounts marked cancelled.
The number became 137.
For six months, Samira’s small team had counted every account that had ever paid, including people who signed up for a pilot, paid once, and left before a second invoice. The number had appeared in decks, grant applications, and a partnership proposal. Nobody had asked the question precisely enough for the definition to surface.
Now somebody would.
The bad ending was straightforward. If the investor found the mismatch first, the meeting could become a conversation about judgment and reporting rather than the product Samira had spent two years building. If she sent 184 without explanation, the larger number would sit in the deck like a loose floorboard.
The count has to match the decision it supports
Customer count sounds simple until you ask what decision it is meant to support.
An investor trying to understand retention needs a current view of paying customers, churned customers, and the period each group belongs to. A sales partner may reasonably want to know how many organisations have ever used the product. A founder deciding whether to hire support staff needs to know who still needs support this month.
Those are different numbers. Trouble starts when one label carries all of them.
Samira had a reason for keeping the old figure. Her company sold a workflow tool to small businesses, and early pilots had been hard won. Each cancellation felt temporary. One customer had paused while changing staff. Another had switched payment methods and never completed the return. A third said they might come back after the holiday season.
That uncertainty is real. It still does not turn a cancelled account into an active customer.
The useful move is to preserve the history without blurring the present. “184 accounts have paid at some point; 137 are currently active” gives an investor a clearer picture than either number alone. It also opens the conversation that matters: why did 47 leave, what patterns appeared, and what has changed since?
Diligence exposes definitions that day-to-day reporting can hide
Most early-stage teams do not build misleading dashboards on purpose. They build quickly, make choices that suit the next meeting, and postpone the cleanup because the product has a more immediate problem.
That can work for a while. Then an external reader asks for the raw export.
The raw export has no patience for deck language. It shows duplicate accounts, paused subscriptions, failed payments, free trials, manually created records, and cancellation dates that nobody had turned into a weekly operating habit. A strong diligence process makes those gaps visible because it compares the summary to the underlying record.
Samira could have sent the old number and hoped the investor focused on the roadmap. Instead, she changed the slide, added a definition under the metric, and prepared a short reconciliation. The report showed active paying accounts, cancelled accounts, and accounts that had paid once but never renewed.
She also wrote down what she did not yet know. Several cancellations had no reason attached because the cancellation flow had been built for speed, not learning. That was uncomfortable. It was also a more credible answer than inventing a clean explanation after midnight.
The same discipline matters when a company is deciding how to spend limited runway. In The Twelve Days of Payroll Kojo Could Not Change With a Funding Headline, the number that matters is the one that can actually cover payroll. Reporting has the same constraint: the useful metric is the one tied to the decision in front of you.
A smaller truthful number can create a better investor conversation
By the time Samira joined the call, she had stopped trying to defend 184.
She led with 137 active customers and explained that the team had previously used a cumulative paid-account count in external materials. She showed the difference, named the reporting change, and described the next action: every cancellation would receive a reason code, with a weekly review of the accounts most likely to leave.
The investor still asked difficult questions. Were the cancellations concentrated among a certain type of customer? Did the team know whether onboarding, pricing, or product reliability drove them? How many active accounts had paid for more than one billing cycle?
Those questions did not disappear because the number was honest. They became answerable.
That is the value of correcting a metric before diligence. You move from protecting a claim to examining the business. A founder can say, “Here is what happened, here is how we now measure it, and here is what we are testing next.” That is a much stronger position than trying to keep every earlier slide intact.
Build the reconciliation before the next high-pressure night
The practical answer is a small metric policy, written before the next investor update.
Define active customer in one sentence. Define churn in one sentence. Keep a separate cumulative count for accounts that have ever paid if it serves a real purpose. Give each dashboard metric an owner and a source. Then compare the dashboard to the export on a regular schedule, especially before fundraising, financing, or a major partnership discussion.
Samira’s next update would still show 137, unless customers actually returned or new ones paid. The difference was that every person in the room would know what the number meant.
At her kitchen table the next evening, the cancellation list was still open. This time, she was not looking for a way to make it disappear. She was writing the questions her team needed to ask before the next account left.
Comments
No comments yet.