Metrics, retention and the organization · Answered by Golden Section from more than 400 B2B software companies observed
Most SaaS companies should hire a controller before a CFO, and hire the CFO when finance work shifts from getting the numbers right to making decisions with them. A controller, in-house or as a service, owns the monthly close, revenue recognition and an ARR schedule that reconciles to the ledger. A CFO earns the seat when the company is raising or refinancing capital, carrying debt with covenants, running a budget the board governs against, or preparing for an exit, and when those decisions are consuming the founder's time. Revenue is context rather than the trigger. The first step is the same either way: get the monthly close landing within fifteen days, because a CFO hired onto unreliable books spends the first two quarters as an expensive controller.
Hire accuracy before strategy. A controller comes first; a full-time CFO comes when capital, covenants, board budgeting or exit preparation are recurring decisions rather than one-off events. The founder keeps responsibility for cash regardless of who holds the title.
| Metric | Value | What it means | Source |
|---|---|---|---|
| Monthly close | within 15 days | days from month end to closed, reviewed financialspart of the financial hygiene a buyer's diligence team expects | Golden Section, publishedWhat Operational Support Actually Means |
| G&A spend, private SaaS median | 15% of ARR | general and administrative cost, including finance, as a share of ARR2026 survey of more than 1,000 companies; public companies run 6% to 9% | External benchmarkSaaS Capital, 2026 Spending Benchmarks for Private B2B SaaS Companies |
| Audited financials | above a $3M raise; worth preparing for from about $2M in sales | third-party audit of annual statementsa review or compilation first builds the auditor relationship | Golden Section playbookAudited Financials play and the $2M investor guide |
The CFO title gets hired too early for a simple reason: the founder feels the pain of the numbers and assumes a senior hire will fix it. But most early finance pain is accuracy, not strategy. Contracts are not in the ARR schedule, revenue recognition is improvised, and the close runs into the next month. Those are controller problems, and a controller or controller-as-a-service fixes them at a fraction of the cost.
The CFO seat pays for itself when the questions change. How much should we raise, and should it be debt? Can we hold this covenant through a slow quarter? What does the budget say about the hiring plan? What will diligence find? Each of those depends on a cash flow forecast that is updated with actuals and a budget the board holds management to. Hiring the CFO too late shows up in a diligence process, where reported and recognized revenue disagree and the discovery costs a turn of multiple. Hiring too early shows up as an expensive executive doing reconciliations.
A company at $7M in annual revenue uses an outside bookkeeper and closes its books around day 35. The founder wants a CFO before arranging a term loan. Instead, the company engages a controller service that brings the close to day 12 within two quarters, reconciles the ARR schedule to the ledger and completes a first review of the financials. With the loan closed and an exit targeted inside three years, the board then approves a CFO search, and the finalist inherits clean books and a working forecast. All figures are illustrative.
A founder with no finance background running a complex business, such as multiple entities, payments flows or significant debt, may need CFO-level judgment earlier; a fractional CFO often fits that gap. A founder with deep finance experience can sometimes carry the CFO role longer, provided a controller owns the close.
From the Golden Section mistakes list, each paired with the play that prevents it.
Hiring finance talent is not the same as handing over responsibility for cash, which stays with the founder.
Books that are reopened after close destroy trust with investors and lenders, and fixing that is the controller's first job.
The cash model has to be updated to current reality every quarter whoever owns it.
An improvised revenue recognition policy is the most common thing a finance hire inherits and the most expensive thing diligence finds.
In the order we would run them. Each is on its own page, most with a free Excel template.
Gives the controller a contract-level ARR record that can reconcile to the ledger.
Builds the auditor relationship through a review before the first audit is required.
Sets up the living cash model that the founder owns early and a CFO takes over later.
Creates the budget with owners and monthly variance reporting that a CFO runs against.
Scores CFO candidates against a written scorecard and a budget set before the search.
Executive plays The executive plays hold the finance operating system: budget, cash forecast, audits, board and hiring.
Large enough to close monthly within fifteen days, maintain the ARR schedule and forecast, run the budget process and support audits and diligence. For most companies that means a controller with accounting support and CFO-level judgment either in-house or fractional. G&A in total runs about 15% of ARR at the private median, and finance is one part of it.
Hire a controller when the problem is accurate books: the close, revenue recognition and reconciliation. Hire a CFO when the problem is decisions: capital strategy, debt and covenants, board budgeting and exit preparation. Most companies need the controller first.
Often, for companies whose CFO-level decisions are periodic, such as an annual budget, a single raise or a loan. It works when a controller already owns the monthly close, so the fractional hours go to judgment rather than cleanup.
Clean monthly financials are the entry ticket for both our products; our lending requires a monthly P&L, ARR bridge and cash flow statement. The finance hire that produces those on time is often the hire that makes capital available.
Growth capital lending →Reviewed by Dougal Cameron, CEO & Co-Founder on 2026-09-23. Golden Section observations are labeled separately from external benchmarks and illustrative arithmetic.