VIII. Your Move: The Role Assignments
Top Takeaways
What the findings mean for an operator, as things to do and things to watch. For the observations behind them, see Key Insights; for the full argument, The Imperative.
The Reframe
What to Do
Four moves, ordered by strength of evidence and how far each one moves the numbers. Each is a decision a company controls directly.
Shift the contract mix back to multi-year
It cuts annual churn from ~14% to ~3% (a 79% reduction), and you set it at the point of sale. Where a 15–20% incentive discount wins the term, it is in most cases a small trade: the discount costs points of margin, while a churned customer forfeits the entire 37-month CAC payback. Adoption fell from 48% to 26% in a year. Two riders: plan the effect at a discount to the headline gradient (part of it is selection and renewal-window timing), and pair the term with takeaway #4 — a multi-year lock on per-seat pricing defers the seat-compression conversation to one leveraged renewal.
Report downsell as its own line
Break it out by cohort, alongside churn and expansion, rather than blending it into NDR. A third of all revenue loss is invisible until you measure it, and downsell re-expands faster than churn, but only if it is seen in time.
Below $25M ARR, grow on new logos, not expansion
Build growth on new-logo efficiency and the retention that makes later expansion possible. Expansion-majority growth is a consequence of reaching scale, not the route to it: the median company never crosses 46%.
Move off pure per-seat pricing
Do it before the 24–36-month inflection. Under a seat model, every headcount reduction an AI tool enables flows straight to revenue with no renegotiation required. The contraction is built into the contract. The hedge is a hybrid: a committed base for predictability, with a usage or outcome component that captures the work AI adds. This and takeaway #1 are one decision made in the same agreement: put the AI-proof value metric inside the multi-year term, or the term converts the seat exposure into a single batched repricing at renewal.
What to Watch
Four instruments that report the truth earlier than the headline numbers do.
Expansion-to-churn coverage, not headline NDR
At 1.07× coverage, a soft quarter that costs a point or two of expansion pushes net retention below 100%. The headline number hides how thin the margin has become.
Fully-loaded CAC against net retention
Reported CAC understates the true cost 2.8×. Cutting R&D or shortening contracts to protect margin erodes the very retention that made acquisition look affordable.
Rule-of-40 achievement, not median EBITDA
A rising median can hide a thinning population: EBITDA improved 48 points while the share of companies clearing Rule of 40 fell from 11% to 5%. Cost cuts buy reported profit while eroding the foundation that reports last.
The slower-reporting metrics are the real signal
Net retention, gross churn, and downsell report last, so improvement in the metric that reports fastest, margin, can finance deterioration in the ones that report slowest. Read them as the true health signal.
From the Deep Dives
Two further takeaways come from the Go-to-Market deep dives, which examine the machinery beneath the eight findings rather than adding to them. They are unnumbered for that reason: the frame above is the findings’; this layer is the machine’s.
Do not plan on selling around a retention gap
Four years of go-to-market iteration left booked ARR per ramped rep at $560K in 2024, essentially its 2020 level, while quotas froze and attainment slid from 75% to 70%. A pipeline plan that assumes acquisition can outgrow weakening retention is planning against a machine already at its limit.
Give Customer Success its own budget line
The benchmark survey has never measured CS spend as a share of revenue, and 57–68% of CS cost sits inside sales and marketing (S&M), so every S&M cut is partly a retention cut no statement shows. Breaking the cost out is what makes a retention investment weighable against the next cut.
Frequently asked questions
What should SaaS operators do about falling retention?
Four moves. Shift contracts back toward multi-year terms, the lever tied to a 79% reduction in churn. Report downsell as its own line. Below $25M ARR, grow on new logos rather than expansion. Move the value metric off the seat.
Why report downsell as a separate line?
Downsell is about 32% of all revenue loss, but both gross and net dollar retention net it away. A company can watch GDR and NDR every month and never see a third of what it is losing.
Are multi-year contracts still worth the discount?
The survey ties multi-year terms to roughly 3% annual churn against 14% on month-to-month agreements, a 79% reduction. Adoption nonetheless fell from 48% to 26% in a single survey year, so the market retreated from its strongest churn lever.
What metric should replace net dollar retention?
No single metric. The report prescribes a decomposed panel separating new, expansion, downsell, and churn, with expansion-to-churn coverage as the one headline figure. A blended number is what let downsell hide for a decade.
Last reviewed: July 2026