Metrics, customers and the organization · Answered by Golden Section from more than 400 B2B software companies observed
Freeze net-new hiring when the cash forecast or the efficiency numbers say the plan cannot fund itself, not because one month disappointed. The clear signals are a forecast that falls below six months of operating expense in reserve with no committed financing, a burn multiple above 1.5x for two quarters (it should sit well below that, ideally under 1x), or sales and marketing spending above $0.70 per dollar of new ARR while you keep adding sellers. A freeze is a pause to rebuild the plan, so give it a review date, honor accepted offers and protect the seats that carry revenue and retention. To grow without adding people, sell expansion to the installed base, reprice the poor-fit customers and put AI against repetitive work. This month, rebuild the budget from initiatives and reopen hiring one seat at a time.
A freeze is a budget decision with an end date. Stop net-new hiring when runway, burn multiple or sales efficiency shows the plan cannot fund itself, then reopen seat by seat, each tied to an initiative with an owner and a measurable payback.
| Metric | Value | What it means | Source |
|---|---|---|---|
| Cash reserve | six months of operating expense at the forecast trough, or an undrawn facility covering the same shock | cash held against a surprise churn or late paymentcash beyond that should have a job; debt costs about 15–20% a year and equity 50% or more | Golden Section, publishedHow much runway should a B2B SaaS company have? |
| Burn multiple bands | under 1x efficient; 1x–1.5x deserves a look; above 1.5x structural | net burn ÷ net new ARR, trailing two to four quartersaim well below 1.5x; sales efficiency, churn or opex misallocated between sales and product usually explains a high reading | Golden Section operating viewGolden Section operating view |
| Sales efficiency target | 0.70 or less | prior-period sales and marketing expense ÷ new ARR booked, including upsellsabout $0.70 of cost per $1 of new ARR; above it, adding sellers adds burn faster than revenue | Golden Section playbookSales Efficiency Ratio |
| Cost of expansion vs new-logo ARR | $0.80 vs $1.63 | sales and marketing cost per dollar of ARRmarket figures; our targets are under $1 for new logos and well under $0.50 for expansion, and the installed base is the cheapest growth during a freeze | Golden Section, publishedInvesting in Software addendum, September 2026 |
| Poor-fit share of a customer cohort | about 20% | customers whose needs or economics do not fit the productraise their price or let them go; either frees capacity without a hire | Golden Section operating viewGolden Section operating view |
Every hire commits cost on the day the offer is signed, while the revenue it is meant to create arrives a sales cycle or a product cycle later, if it arrives. When the cash flow forecast shows the reserve being consumed, or the burn multiple has stayed in the structural zone, more hiring is scaling a system that is not yet converting. A freeze buys the time to find out why.
But a blanket freeze is a blunt tool. Stopping a support backfill can cost more in churn than the salary saves, and strong people read an unexplained freeze as a leadership team that does not know what is wrong. So rebuild the budget from initiatives, keep the seats that carry renewals and implementation, and cut the open requisitions nobody can tie to a measurable payback. Check sales efficiency by channel before you add a seller back.
Revenue without headcount comes mostly from the installed base. Account management that pitches expansion converts at about half the cost of a new logo, and the poor-fit fifth of customers is usually underpriced for the effort it takes. AI changes the hiring plan the same way: before each requisition, ask whether the work is repetitive enough that a tool absorbs it.
The numbers are invented. A company at $3M in annual revenue has run a burn multiple near 2x for two quarters, and the forecast shows its reserve dropping below six months of operating expense by spring. It has five open requisitions. It honors one accepted offer, backfills one support role, and pauses three sales hires and a marketing hire with a 60-day review date. Sales efficiency by channel shows one channel near 0.6 and the rest above 1.2. The company pitches an expansion module to its 40 largest accounts and reprices the poor-fit fifth at renewal. Two quarters later the burn multiple is near 1.3x, still above the 1x it is aiming for, and it reopens one account executive seat, assigned to the efficient channel.
A company that is profitable and burning nothing does not need a freeze to protect runway, but it may still need one to stop hiring that has run ahead of revenue per employee. And burning to grow is acceptable while sales efficiency stays at or below about 0.70, the burn multiple sits well below 1.5x, and sales and marketing spend accounts for the burn; the freeze belongs on the seats outside that engine.
From the Golden Section mistakes list, each paired with the play that prevents it.
A freeze decided late, because nobody was watching burn and runway monthly, has far fewer good options.
Hiring into a structural burn multiple scales a broken engine instead of fixing it.
Every team will say a freeze leaves it too thinly staffed, and that complaint cannot be the test for reopening a seat.
In the order we would run them. Each is on its own page, most with a free Excel template.
Shows when the reserve is consumed, which is the trigger that matters most.
Separates the channels worth hiring into from the ones adding burn.
Rebuilds the plan from initiatives so every reopened seat has an owner and a reason.
Turns the installed base into expansion revenue that does not need new hires.
Executive plays Budget, cash and operating plays that decide how much headcount a company can afford and which seats come first.
Start with the installed base, where a dollar of expansion costs about half what a new-logo dollar does. Reprice or exit the poor-fit fifth of customers, check whether customers earning more than ten times their cost are underpriced, and put AI against repetitive support and research work. Then measure revenue per employee by department to see which of these is actually working.
Ask of every requisition whether the work is repetitive and written enough that a tool could absorb it, and if so, test that before hiring. Expect fewer hires per dollar of new revenue in support, implementation and sales development, and plan the savings into product and the inference line rather than assuming them before they are measured.
Not for seats that carry revenue or retention, such as support, implementation and quota-carrying sellers on an efficient channel. Leaving those empty costs more in churn and missed bookings than the salary saves. Treat every other backfill as a new requisition that has to justify itself.
A freeze to protect runway is usually a fix-first moment, and capital raised to avoid one tends to fund the problem. Once a channel has proven its efficiency again, non-dilutive debt can fund it without new dilution.
Growth capital lending →Reviewed by Dougal Cameron, CEO & Co-Founder on 2026-10-22. Golden Section observations are labeled separately from external benchmarks and illustrative arithmetic.