Capital, cash and fundraising · Answered by Golden Section from more than 400 B2B software companies observed
In our experience across thousands of board meetings and founder engagements, the biggest lever is sharper qualification and better account management, not a line-by-line cut. Selling what you have to customers with exactly that problem makes new-account growth cheaper, and working the installed base surfaces revenue you have already earned the right to. That focus lets you trim product spend: activate AI productivity across the team, then cut the department by about 20% while holding the same output goal. Next, find the roughly 20% of customers in any cohort who got through loose qualification, and ask them for a price increase or help them leave. Protect sellers on proven channels, implementers and support, and stop any channel costing more than $1.00 per $1 of new ARR. Start by scoring every customer on fit.
Grow by qualifying tighter and managing accounts deeper, then fund it by trimming product spend and repricing or exiting poor-fit customers. Protect proven channels, onboarding and support, and make any headcount reduction a single decision sized from the forecast.
| Metric | Value | What it means | Source |
|---|---|---|---|
| Sales efficiency threshold | $1.00 per $1 of new ARR | sales and marketing expense ÷ new ARR bookedworse than this, fix the motion before financing it | Golden Section, publishedGrowth Capital Without Heavy Dilution |
| Cost of expansion vs new-logo ARR | about $1.00 vs $2.00 median | sales and marketing cost per $1 of ARRprivate B2B SaaS, as cited in our September 2026 addendum | Golden Section, publishedInvesting in Software |
| Sales and marketing as % of revenue | 33% PE-backed vs 47% venture-backed | S&M expense ÷ revenueprivate SaaS at comparable scale; ownership model changes the spend, not the business | Golden Section, publishedInvesting in Software |
| Non-discretionary cost lines | support 9%, G&A 15%, R&D 22% of revenue | median spend by functionSaaS Capital 2026 spending survey of 1,000+ private B2B SaaS companies, as cited in our addendum | External benchmarkSaaS Capital 2026 spending survey, cited in Investing in Software |
| Poor-fit customers | about 20% | customers who got through loose qualification and are not a real fitin any customer cohort; they should pay more or leave | Golden Section operating viewGolden Section operating view |
| Product spend reduction | about 20% | reduction in product department cost while holding the same output goalafter AI productivity is activated across the team | Golden Section operating viewGolden Section operating view |
Burn falls fastest when you cut the largest line, which is almost always people, so founders reach for layoffs first. But much of the burn at this stage traces back to loose qualification: sellers chasing long shots that will never be core customers without radical product changes, and a product team building for them. Qualification rooted in selling what you have to people with that exact problem lowers the cost of new ARR and shrinks the roadmap at the same time. Customer segmentation tells you who those people are, so you can go deeper with current customers and find more that look exactly like them.
Account management is the other growth line. Expansion ARR costs roughly half what a new logo does at the median, and the same process finds the misfits. There are always about 20% of customers in a cohort who are not a real fit; they burn out your team and are not worth the revenue. Demand an increase or show them the door. It is a small step back that frees capacity as the sharper focus takes hold.
Then do the unglamorous work in the vendor contract register and receivables. If the forecast still falls short of self-funding, one well-sized reduction does less damage than three small ones.
A company at $4M ARR burns $200K a month and needs to cut $80K without slowing growth. Its customer fit scores show the best accounts are mid-sized firms in one sub-segment, yet half of last year's pipeline sat outside it. It narrows qualification to that profile and drops a trade show program booking at $2.40 per dollar of new ARR, saving $35K a month. With AI tools activated, product holds its output goal with 20% fewer people, saving $30K. Of 120 customers, 24 are poor fits; 15 accept a price increase and 9 leave over two quarters, a small ARR dip that frees support time for expansion work with the best accounts. Renegotiated contracts cover the last $15K. No seller on a proven channel is cut. The figures are invented.
If the proven channels are themselves inefficient, above $1.00 per dollar of new ARR, protecting them protects a leak, and the fix is the sales motion rather than the budget. When cash is already inside the reserve, speed matters more than precision.
From the Golden Section mistakes list, each paired with the play that prevents it.
Loose qualification fills the pipeline with long shots and drives product spend toward customers who will never be core.
The poor-fit 20% of customers burn out the team; holding them at any price costs more than the revenue.
Delaying the cost decision makes the eventual cut larger and the options fewer.
If headcount has to go, a clear and decisive termination does less damage than a slow one.
In the order we would run them. Each is on its own page, most with a free Excel template.
Defines the best-fit customers that sharper qualification and lookalike prospecting depend on.
Finds expansion in the installed base and gives poor-fit customers an orderly offboarding.
Ranks channels by what they cost per dollar of new ARR.
Surfaces renewals and notice dates so unwanted spend can be stopped on time.
Tells you whether the cuts reach self-funding, and whether headcount must change.
Executive plays Cost and cash decisions sit with the founder, and the executive plays cover the forecast and budget that frame them.
Go deeper with the customers you have and find more that look exactly like them. Expansion ARR costs about half what new-logo ARR does at the median, and tighter qualification makes new logos cheaper too. Fund it by trimming product spend once AI productivity is in place and by cutting your least efficient channel.
Long-shot deals and the product work built to win them, then channels with the worst cost per dollar of new ARR, then product headcount once AI tools are activated, holding the same output goal. Next come poor-fit customers, who should pay more or leave. Leave onboarding, support and your best sellers for last, and do not spend weeks on small line items.
When the monthly forecast, after every other cut, still does not reach self-funding or a closed financing with six months of operating expense in reserve. Then size the reduction to the forecast, do it once, and communicate it clearly. Repeated small cuts cost more in trust and productivity than one decision made early.
Yes. About 20% of customers in any cohort are not a real fit, and they consume more support and roadmap than their revenue justifies. Ask for an increase that makes them worth serving, and offboard the ones who decline. The revenue dip is small and the capacity it frees goes to your best accounts.
When an efficient channel is short of cash rather than failing, debt can fund it while you cut elsewhere; our lending funds channels with CAC payback under 18 months. When the channel itself is inefficient, capital is the wrong fix.
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.