SaaS CAC Benchmarks: What Customer Acquisition Actually Costs, From Real Disclosures

Customer acquisition cost is the fully loaded price of winning one new customer:

CAC = total sales + marketing spend in a period ÷ new customers added in that period

“Fully loaded” is where most self-reported numbers quietly cheat: the honest version includes salaries, commissions, tools, content, events and management overhead — not just the ad budget. The benchmark question every founder asks — what should CAC be? — has no single answer, because CAC is only meaningful against what a customer is worth and how fast the money comes back. What follows are the real numbers founders have disclosed on the record, organized the way operators actually think: by sales motion.

Self-serve: CAC in double digits

At sub-$1K contract values, acquisition must be nearly free, which means product, content and community do the selling — and the “spend” is really the years of compounding investment behind the organic channel. That’s the pattern in the disclosed numbers:

Mid-market: the 12–18 month window

Where inside sales enters, disclosed paybacks cluster tightly. Salesloft’s Kyle Porter: 13–14 months at a $50M run rate. Algolia blended: under 12 months. Showpad’s Pieterjan Bouten: closer to two years blended, justified by an LTV:CAC ratio of about 4 and single-digit gross churn — his ~$200K lifetime values could absorb it. The working benchmark from these disclosures: a mid-market SaaS motion is healthy when payback lands under 18 months and strained past 24.

Enterprise: big absolute CAC, engineered payback

Workboard’s Deidre Paknad disclosed a $17–18K CAC on $125K first-year contracts — and then the move that made the number almost irrelevant: customers paid for coaching-led onboarding as a services fee, which repaid acquisition cost essentially on day one while staying out of the ACV line. The full mechanics are here. Enterprise CAC in the tens of thousands is normal; what separates operators is how deliberately they finance it.

The rule the best operators actually run

Outreach’s Manny Medina gave the cleanest formulation in the archive, and it’s stricter than the usual folklore: every segment must clear gross-margin-adjusted CAC payback under 20 months. The adjustment matters — recovering CAC from revenue overstates speed by whatever your margin isn’t. At his ~80% gross margin, a nominal 16-month payback is really 20. He ran the same efficient-frontier logic per segment: “figure out the right ACV, CAC and deal-length combination for that portfolio.” The related ratio worth tracking once you’re past $1M ARR: new ARR added per dollar of S&M spend (the “magic number”), where disclosed-healthy looks like Workboard’s $2.50 of new revenue per $1 spent — deliberately relaxed from $5 as she bought growth.

Reading your own number honestly

Four disciplines separate a real CAC from a slide CAC:

• Load everything — salaries and overhead, not just media.
• Segment it — Algolia’s blended sub-12-months averaged near-zero self-serve with “pretty high” enterprise, and only the split is actionable.
• Adjust for margin — per Medina.
• Pair it with retention before judging it — Showpad’s two-year payback was fine because churn was single-digit; the same payback with 20% revenue churn is a slow-motion write-off.

CAC is never good or bad alone — it’s one leg of a triangle with contract value and retention, and the triangle either closes or it doesn’t.

Compare disclosed CAC, payback and retention across thousands of SaaS companies at getlatka.com/saas-companies.