Metrics, retention and the organization · Answered by Golden Section from more than 400 B2B software companies observed
Allocate it through the budget, in a fixed order. Fund product and engineering to what the roadmap and the installed base require, because underfunding product is the slow way to lose a vertical market. Add sales and marketing capacity only as fast as sales efficiency proves it converts, and hold G&A and customer success at the level the business needs to close its books and keep customers. As a reference, private B2B SaaS companies spend a median 22% of ARR on R&D, 23% on sales and marketing, 15% on G&A and 9% on support and success, and people are most of each line. Use those as a check on the plan rather than a template, and let each department's hires trace to a strategic initiative with an owner.
Headcount follows strategic initiatives through the budget, not percentages. Sales and marketing capacity scales with demonstrated sales efficiency; product capacity scales with the roadmap and the base it protects; G&A stays near the floor required to run the company well.
| Metric | Value | What it means | Source |
|---|---|---|---|
| Spend by function, private B2B SaaS median | R&D 22%, sales 15%, marketing 8%, support and success 9%, G&A 15%, hosting 5% | share of ARR2026 survey of more than 1,000 companies; at $3M to $5M ARR, R&D 24%, sales 12%, support 10% | External benchmarkSaaS Capital, 2026 Spending Benchmarks for Private B2B SaaS Companies |
| Sales and marketing, PE-backed vs venture-backed | 33% vs 47% of revenue | sales and marketing expense as a share of revenue at comparable scaleownership model changes the spend more than the business | Golden Section, publishedInvesting in Software, 2026 addendum |
| Engineering and product allocation in the AI era | about 45% core, 45% AI-native features, 10% skunkworks | share of engineering and product resourcesGolden Section's suggested split for incumbent software companies | Golden Section, publishedThe Enhancement Doctrine, March 2026 |
| Sales efficiency ratio target | near 0.7 | prior-period S&M expense per $1 of new ARR bookedabove 0.7 is inefficient spend; well below suggests room to spend more | Golden Section playbookSales Efficiency Ratio |
Departments compete for headcount with the arguments they have, and sales usually has the most direct one. The failure is predictable: sellers get hired ahead of a motion that converts, product stays underfunded while the installed base asks for more, and G&A is either starved until the books cannot close or grows because nobody owns it. A budget with an owner and accountable lead on every department line is the mechanism that stops this, because each hire has to trace to an initiative and survive a review where every other leader is looking at it.
The sales efficiency ratio is the gate for sales and marketing hires. It tells you whether the next dollar of spend creates ARR at a rate worth buying. For product, the question in 2026 is also what the engineers are building. The Enhancement Doctrine argues for roughly equal effort on protecting the core and building AI into existing workflows, with a small team on products that could replace the core. Engineering headcount without that split tends to drift entirely toward one side.
A company at $4M ARR has 32 people: 9 in engineering and product, 12 in sales and marketing, 6 in implementation and support, and 5 in G&A. Its sales efficiency ratio has drifted from 0.8 to 1.3 over four quarters while the roadmap slips. The budget review freezes two open sales seats, moves the funds to two engineers assigned to the workflow feature customers keep asking for at renewal, and adds a controller so the monthly close lands inside fifteen days. Total headcount stays at 32. All figures are illustrative.
Companies with heavy implementation or a services line will carry more people in delivery than the benchmark suggests, and that is fine when services margins and retention justify it. A company entering a second vertical may run sales above benchmark for a planned period.
From the Golden Section mistakes list, each paired with the play that prevents it.
Product is the department most often shorted when sales pressure sets the allocation.
Adding sales headcount to a motion with poor efficiency scales the problem instead of the revenue.
Allocation means nothing if overlapping roles leave no single owner for each core activity.
In the order we would run them. Each is on its own page, most with a free Excel template.
Runs the allocation through owners, a defended review and monthly variance reporting.
Gives a single number that decides whether the next sales and marketing hire is worth funding.
Structures engineering roles and reporting lines so product headcount maps to accountability.
Chooses the sales structure so new sales seats fit a model rather than accumulate.
Executive plays The executive plays own the budget, the operating system and hiring, which together decide where headcount goes.
The private median is 22% of ARR, and 24% in the $3M to $5M band, per SaaS Capital's 2026 survey. Vertical software protecting a workflow and building AI into it usually should not run far below that. Underfunding product is one of the mistakes we see cost the most over time.
There is no published headcount ratio we would stand behind, and headcount mix depends on how much implementation work you carry. Since R&D is about 22% of spend at the median and people are most of that line, a company with well under a fifth of its team in engineering and product should be able to explain why.
Median private spend is about 23% of ARR, 15% on sales and 8% on marketing. PE-backed companies run sales and marketing near 33% of revenue against 47% for venture-backed peers at comparable scale. The right level is whatever your sales efficiency justifies, which is why we gate sales hires on that ratio.
About 15% of ARR at the private median, falling toward 6% to 9% at public-company scale. G&A should be sized to close the books monthly, run the budget and keep the company diligence-ready, and not beyond it.
Start from the strategic initiatives for the year, assign each to a department owner, and have each owner defend the hires the initiative requires in a budget review. Then check the result against revenue per employee and spend-by-function benchmarks, and revisit the plan monthly against actuals.
Allocation decides what capital is for. A proven sales motion with repeatable efficiency can often be funded with non-dilutive debt; a shift in product investment, such as building AI into the core workflow, is a structural change better suited to equity.
Talk to Golden Section →Reviewed by Dougal Cameron, CEO & Co-Founder on 2026-09-23. Golden Section observations are labeled separately from external benchmarks and illustrative arithmetic.