Find where the money is going missing

When to run this

Run this when a healthcare business has real, verifiable activity — patients seen, orders placed, referrals flowing — but the cash landing in the account does not match it, and nobody can point to the specific step where the value goes missing. It is built for a founder or finance lead who has two working days, read access to their own systems, and a willingness to look at unflattering numbers before a board meeting or investor update forces the issue.

It is not a financial statement audit, and it will not produce an audited number. It will not fix a business with no real demand — if the top of the funnel is empty, this tool will just tell you that quickly. And it is the wrong instrument for a slow structural problem like mispriced products or a broken value proposition; those need the 5-Phase Commercial Turnaround Framework, not a 48-hour sprint.

What to request before the clock starts

Ask for read-only exports or screenshots, fixed to one cut-off date. Do not let anyone "clean up" the source files before you see them — a discrepancy someone tidies away before the audit is the discrepancy the audit exists to find.

The hour-by-hour runbook

The first two hours are the ones teams try to skip, and they are the ones that determine whether the rest of the audit means anything. If "delivered" means something different to operations and to finance, every number downstream is an argument waiting to happen. Write the definitions down before touching a spreadsheet.

The leakage map

Test each control point in order. A control point cannot be marked clean on the strength of someone's confidence in it — it needs a result and a source.

Core calculations

Hold the period and unit of measure constant across all three lines, or the waterfall will not close.

Gross expected value = delivered units × approved price
Invoiced value       = valid invoices/claims linked to delivery
Collected value       = receipts linked to valid invoices/claims

Delivery leakage     = gross expected value − invoiced value
Collection leakage   = invoiced value − collected value
Unverified exposure  = partner claims without agreed evidence

Give every item a unique ID and exactly one primary leakage reason. An item that fails at both billing and collection gets filed once, at whichever step it failed first — otherwise the same naira gets counted twice and the total becomes fiction.

Evidence grading

This is the part of the audit most teams are tempted to skip, because it forces a distinction between what is true and what someone is fairly sure of.

How to read the result

The useful output of this audit is not a dramatic leakage figure. A large number with no owner attached is a press release, not a control. The output that matters is a short list — ideally under ten lines — of specific leakage mechanisms, each with a source record and a name against it.

Two failure modes to watch for, both from inside the audit rather than outside it. First, teams round Grade C evidence up to B because the story is plausible and everyone is tired by hour thirty — resist this, a plausible story is exactly what unverified exposure looks like. Second, teams use the audit to assign blame rather than next steps, which guarantees the next audit gets less honest data, because whoever owns the weak control point will manage the appearance of the numbers rather than the numbers themselves.

Seven-day action log

Every line needs a proof-of-closure field filled in before it moves off "open." "Followed up" is not proof. A corrected invoice number, a reconciled receipt, or a signed delivery note is proof.

Worked walkthrough

Take a generic example: a diagnostics company running twelve pharmacy partners in Ikeja as a sample-collection and results channel. Monthly test volume looks healthy on the pharmacy activity reports, but collected revenue has trailed reported activity for three months running, and nobody in the leadership team can say by how much or why.

The 0–2h window fixes a definition that turns out to be contested: "delivered" has meant "sample collected" to the pharmacies and "result issued" to the lab team, a gap of one to four days in which a sample can be lost, rejected, or re-run without anyone recording it against the original order ID. That single definitional gap, once named, already explains a visible slice of the discrepancy before a single spreadsheet is opened.

The reconciliation pass (2–8h) matches pharmacy-reported activity against lab-system order IDs and finds roughly one in eight pharmacy-claimed tests has no matching lab record — Grade D, so it does not count as leakage yet, only as unverified exposure needing a follow-up call to each pharmacy lead. The price-and-collection pass (8–16h) finds three pharmacies applying a discount approved for one partner only, never updated in the shared price sheet — Grade A collection leakage the moment the ledger is checked against the invoice. By hour forty, the exception list has nine partner discrepancies, three financially quantified and six flagged for a defined evidence check rather than a guess. That is a usable readout for a Monday leadership meeting: a short list of named, ownable items, not a headline percentage nobody can defend.

If you need a second pair of eyes on a leakage number before it goes to a board or investor, contact Dr Tolu Ajidahun.