7 min read · 1,321 words
At a glance
- The annual audit looks at what happened: material misstatements, internal controls, going concern. A Diagnostic looks at what’s about to happen. Different instruments, different jobs, and both worth doing.
- The Diagnostic scans five places, the five canon categories: Cost, Pricing, Growth Spend, Product Line, Capital Structure. Each one holds a class of recoverable value no audit is designed to surface.
- Annual audits cost more. Diagnostics surface more. Only one of them defends the next call.
A Diagnostic finds recoverable value in five places: Cost, Pricing, Growth Spend, Product Line, and Capital Structure. An annual audit looks at none of them, and it is not supposed to. The audit looks at what happened. Material misstatements. Internal controls. Going concern. A Diagnostic looks at what’s about to happen. The two instruments share a building and almost nothing else.
This matters because a lot of post-Series A founders treat the clean audit opinion as evidence the finance function is doing its job. It is evidence of something real, but not that. The audit certifies that the past is accurately recorded. It says nothing about whether the next call is right, and the next call is where companies stall.
What is the annual audit actually for?
Assurance, for people outside the building.
The audit exists so that investors, lenders, acquirers, and regulators can rely on the financial statements without re-deriving them. The auditor tests whether the statements are free of material misstatement, whether the controls that produced them are sound, and whether the company can plausibly continue as a going concern. The work is backward-looking by design: you cannot attest to a number that has not happened yet.
That is the right job, done the right way. When the bank asks for audited financials before extending the line, when the Series C lead’s counsel opens diligence, when an acquirer’s QoE team shows up, the audit is the document that lets those conversations move. We would never tell a venture-backed company past $3M ARR to skip it. The point of this post is not that audits are the wrong instrument. It is that they are an instrument pointed in one direction, and the stall lives in the other one.
Why do audit-clean companies still stall?
Because the stall is not an accounting failure. It is a decision failure, and the audit does not test decisions.
Every line item we are about to walk through below would survive an audit untouched. The $8,000-a-month contract is properly recorded, properly accrued, properly disclosed. The auditor confirms the expense is real. Nobody in the engagement is asked whether the expense is still right. “Right” is a forward question, a judgment about what the company should do next, and forward questions are explicitly outside audit scope. The opinion letter says the books are accurate. Accurate books can describe a company making poor calls in perfect detail.
This is the pattern we see in stalled portcos with clean opinions three years running: the recording function works, the deciding function has quietly stopped, and everyone is reading the wrong instrument for reassurance. The metrics version of this failure has its own write-up in Why retrospective finance fails growth-stage companies.
Where does the Diagnostic look?
Five places, five canon categories. Here is each one, with a worked example of the kind we put in dollars during the scan.
- Cost. The recurring lines over $5K a month that haven’t been re-priced. Worked example: an observability contract at $8,200 a month, signed two years ago when the engineering org was 60 people. The org is now 34, the usage tier is wrong by half, and the vendor’s own pricing page would quote the current footprint at roughly $3,700. Nobody re-opened it because the contract auto-renewed and the line was “on plan.” Re-pricing recovers about $54,000 a year, and the audit would have confirmed the $8,200 was recorded correctly every single month.
- Pricing. The active pricing decisions sitting in a draft because nobody owns the call. Worked example: a 12% list-price increase, modeled in October, supported by a win-rate that has not dipped below 38% in three quarters. The doc is finished. The decision has no owner and no date, so it has rolled forward through two planning cycles. Against a $1.8M renewal base, the unsigned draft is holding roughly $216,000 of annual revenue in suspension. The Diagnostic’s job is not to redo the analysis. It is to put the call on a date.
- Growth Spend. The channel that converts but the unit economics no longer support. Worked example: a paid channel spending $35,000 a month that still converts at the same CAC it did at the last raise. The problem is the denominator moved: gross margin on the acquired cohort fell from 74% to 61% as support load grew, so payback stretched from 14 months to 26. The channel dashboard is green because it tracks conversion, not payback. The spend keeps flowing because the metric everyone watches keeps passing.
- Product Line. The line costing more in support than it returns in margin. Worked example: a legacy SMB tier generating $340,000 of ARR that consumes 40% of all support tickets and half of one senior engineer. Fully loaded, the line returns negative $90,000 a year, and it has survived four planning cycles because $2M of past engineering went into it and killing it feels like writing that off. That feeling is Sunk Cost, and it is mapped in The trap-stack inventory.
- Capital Structure. The multi-year commitment that should be a sublet. Worked example: a five-year office lease signed for 80 seats during the last raise’s hiring plan. Badge data shows 25 people in the building on the busiest day. Subletting two floors recovers roughly $240,000 over the remaining term. The lease is properly disclosed in the audit’s commitments footnote, which is exactly the point: the audit discloses the commitment, and the Diagnostic asks whether the commitment should still exist.
Notice what every example has in common. The books were right. The decision was stale.
What does the 14-day scan actually look like?
Days 1 to 4 are the data pull: GL, contracts, payroll, pipeline, cohort revenue, and the commitments register. Days 5 to 10 run the five categories against that data and score the Six Traps against the actual financials, because every stale decision above is being held in place by a named bias, not by bad math. Days 11 to 14 turn the findings into dollars and dates: each item gets a recovery value, an owner, and a forward decision it unblocks. The full day-by-day walk-through is in The first 14 days: the operational scan.
The Diagnostic is priced from $10,000 to $20,000. A first-year audit for a company that size typically runs $25,000 to $60,000. Annual audits cost more. Diagnostics surface more. Different work, both worth doing. But only one of them defends the next call.
What is the Decisive Finance role in this?
The Diagnostic is ours: a 14-day scan across the five canon categories, run against your actual financials, ending in a dollar figure and a decision list rather than an opinion letter. Guaranteed 3x your Diagnostic fee in recoverable value, in 14 days. Typical 5x to 10x. Keep your auditor; they are doing a job we do not do. We don’t audit the past. We justify each next call.
Where to go from here
If the last audit came back clean and the company still feels stuck, the books are not the problem, and the next instrument to run is the forward-looking one.
Read the guarantee Book the call
The fastest way to see the difference is the deliverable itself. Preview the Trapped Value Report in four minutes.
Related reading: