What is CAC payback and why your board asks for it before ARR
CAC payback period measures how many months it takes a new customer to generate enough gross margin to cover the cost of acquiring them. It is not ARPU, not LTV, not MRR growth: it is the answer to one simple question — does the dollar I spent on sales and marketing come back before the customer churns?
Series B and C funds in 2025 review CAC payback before ARR growth. Two companies with the same ARR and the same growth rate can have diametrically opposite efficiency profiles: one burns 1.20 USD per new dollar of ARR with a 14-month payback; the other burns 3.40 USD with a 38-month payback. The first one accelerates, the second one runs a bridge round.
Formula and numeric example
CAC = S&M spend in the period ÷ New customers in the period CAC Payback (months) = CAC ÷ (Monthly ARPU × Gross margin %)
Example — Velmora Cloud, Series A vertical SaaS. Q3: total S&M spend 540,000 USD, 72 new customers signed. CAC = 540,000 ÷ 72 = 7,500 USD. Average monthly ARPU 620 USD. Gross margin 78% (hosted SaaS, cloud included). CAC Payback = 7,500 ÷ (620 × 0.78) = 7,500 ÷ 483.6 = 15.5 months. Below the 2025 median band for B2B mid-market SaaS — a Magic Number sustained above 0.9 justifies a 30-40% increase in the Q4 S&M budget.
SaaS Magic Number: the quarterly version of CAC payback
Coined in Salesforce's S-1 and popularized by Scale Venture Partners, the Magic Number annualizes quarterly efficiency.
Magic Number = (Quarterly new MRR × 4) ÷ Previous-quarter S&M spend Or equivalent: ((Current Q ARR − Previous Q ARR) ÷ Previous Q S&M spend)
Direct reading: >1.0 the machine is ready to step on the gas; 0.5-1.0 healthy range, invest carefully; <0.5 the engine is broken, cut spend before growing. The one-quarter lag exists because one period's S&M converts into the next period's revenue — not measuring it that way inflates the number 20-30%.
CAC payback SaaS 2025 benchmarks by stage and segment
| Stage / Segment | Median | Good | World-class |
|---|---|---|---|
| Seed SMB SaaS | 14 months | <10 | <7 |
| Series A mid-market SaaS | 18 months | <14 | <10 |
| Series B/C enterprise SaaS | 24 months | <18 | <12 |
| PLG self-serve | 9 months | <6 | <3 |
| Outbound sales-led enterprise | 28 months | <22 | <15 |
Source: OpenView SaaS Benchmarks 2024, ChartMogul SaaS Metrics 2025, Bessemer State of the Cloud 2025. The global B2B SaaS median closed 2025 at ~19 months, up from ~15 months in 2021 — acquisition cost climbed faster than pricing.
Blended CAC vs CAC by channel: the number that fools the board
Blended CAC (all customers over all S&M) is the number that ends up in the board deck. It is useful as a headline, useless for decisions. Breaking it down by channel changes the story:
- Organic inbound / SEO / content. CAC typically 30-50% of blended. Payback 6-10 months.
- Paid ads (Google, LinkedIn, Meta). CAC 1.5-2.5x blended. Payback 18-30 months in B2B SaaS 2025.
- Outbound SDR. CAC depends on ACV; in mid-market with ACV >25K USD it is profitable, in SMB it almost always destroys value.
- Partnerships and referrals. CAC 20-40% of blended. Payback 4-8 months — the most under-invested channel in most startups.
If blended looks acceptable but paid is carrying 60% of the volume with a 28-month payback, the company is growing by buying contracts that never pay back. Killing that channel and reallocating to inbound or partnerships cuts gross growth 15-20% but lifts the Magic Number from 0.6 to 1.1 — exactly the conversation that moves a Series B from 'watch' to 'term sheet.'
Contrarian: CAC payback under 12 months is overrated when NRR exceeds 130%
The obsession with short payback is orthodox pre-2022 SaaS thinking, when capital was cheap and contracts were monthly. In a business with net revenue retention above 130% (Snowflake, Datadog, MongoDB territory), a 24-30 month CAC payback is rational: the customer expands so much in years 2-4 that LTV:CAC rises to 5-8x even though the initial payback is long. Forcing payback to be short by pushing self-serve contracts with 100% NRR optimizes the wrong metric. Real rule: payback matters relative to NRR and gross margin, not in absolute terms.
Mistakes that kill efficiency (and the pitch for the round)
- CAC without gross-margin adjustment. Using gross ARPU inflates payback 20-25%. A SaaS with 78% GM is very different from one with 45% (services heavy). The board adjusts for this, and so should you.
- Forgetting fully-loaded salaries in S&M. Salaries + payroll taxes + equity + tooling (Outreach, Salesforce, 6sense, Gong). In mid-market SaaS, tooling and equity represent 25-35% of an AE's cost, not 10%.
- Mixing acquisition cohorts in the Magic Number. A quarter with a large upsell inflates new MRR and hides a stalled new-business engine. Segment it: new-logo Magic Number vs expansion Magic Number.
- Not measuring against contract cycle. On annual contracts, a 14-month payback means the first customer paid back before renewal. On monthly, 14 months requires churn below 2% to be profitable — arithmetic, not opinion.
- Comparing against 2020 benchmarks. Capital was 5x cheaper and CAC 2-3x lower. Useful benchmarks today are 2024-2025 (OpenView, ChartMogul, Bessemer, Battery Ventures).
When to use this simulator and when not
Use it when you have at least 90 days of clean data: fully-loaded S&M spend, new customers per period, average ARPU per cohort, and audited gross margin. Use it quarterly for board reporting and monthly if you are in an active fundraise. Do not use it pre-revenue (the CAC for the first 20 customers is almost always founder-led and predicts nothing) nor with fewer than 30 active customers. It also does not replace a retention cohort model — payback assumes constant churn, which in reality varies by channel and ACV band.
Cross-link: metrics that complete the story
CAC payback is an input metric. To read the full business you need: [Churn and LTV](/simulador/saas/churn-ltv) to validate the customer lasts long enough to justify payback; [CAC vs LTV](/simulador/saas/cac-vs-ltv) for the aggregate ratio; [MRR Projection](/simulador/saas/mrr-proyeccion) to tie the Magic Number to the revenue forecast; [Ad Campaign ROI](/simulador/marketing/roi-campanas-ads) to attribute channel by channel; and [Subscription Pricing](/simulador/saas/pricing-suscripcion) when payback can only be fixed by raising price.
Industry benchmarks 2026
| Segment | Median payback | Good | World-class |
|---|---|---|---|
| SMB SaaS | 5-9 months | <5 | <3 |
| Mid-market SaaS | 12-18 months | <12 | <8 |
| Enterprise SaaS | 18-30 months | <18 | <12 |
| PLG self-serve | 4-7 months | <4 | <2 |
Source: OpenView SaaS Benchmarks 2025, ChartMogul Annual Report 2026. SMB payback tightened in 2025-2026 as AI-native competitors compressed sales cycles; enterprise payback extended slightly as procurement scrutiny increased post-2023 budget freezes.
Common mistakes that inflate payback
- Counting only paid CAC. Including only ad spend and ignoring organic-inbound infrastructure cost (SEO team, content, tooling) understates blended CAC by 30-50% in mature growth organizations. The board will ask; have the fully-loaded number ready.
- Ignoring activation cost. Customer success onboarding, implementation support, and integration engineering are variable costs that run 8-22% of first-year ACV in mid-market SaaS — but are rarely included in the CAC denominator. Omitting them inflates reported payback by the same proportion.
- Treating gross margin as 100%. Using ARPU instead of ARPU × gross margin to compute payback is the single most common error in early-stage SaaS boards. A company with 60% gross margin (services-heavy, proprietary infrastructure, heavy support) has a real payback 40% longer than the one reported without the adjustment.