CAC explained simply
Contents:
What CAC actually measures
Customer Acquisition Cost (CAC) is the all-in price of turning a stranger into a paying customer. If you spent $100,000 in a quarter to land 400 new paying users, your CAC is $250. Simple division — and yet on most growth dashboards, three different teams will quote three different numbers for "CAC" and all of them will be technically defensible. That gap is where bad unit economics hide.
CAC is one of the three load-bearing metrics any subscription or transactional business runs on, alongside LTV and churn. Investors at Stripe, a16z, and Sequoia don't really care about your raw CAC — they care about the LTV/CAC ratio and the CAC payback period. If LTV/CAC drops below 3 or payback stretches past 24 months, the company is either burning runway, mispricing, or growing on borrowed thesis.
This piece is the conceptual version — the why and the what. If you want the SQL recipes, how to calculate CAC in SQL and how to calculate blended CAC in SQL cover the queries.
Blended, paid, and fully-loaded CAC
Three flavors of CAC show up in every board deck. They answer different questions and you should never use them interchangeably.
Blended CAC divides total acquisition spend by every new customer — paid and organic. It's the version investors pressure-test because it's the hardest to game. If your blended CAC drifts up while paid CAC stays flat, you're losing organic leverage. That's usually the leading indicator nobody watches.
Paid CAC divides paid-channel spend by customers attributed to paid channels. It's the operating metric for performance marketing — the lever growth teams actually pull weekly. Paid CAC will always be higher than blended CAC, sometimes by 3-5x in a brand-driven business like Notion or Figma.
Fully-loaded CAC adds the salaries of the growth team, tooling (HubSpot, Segment, Amplitude, ad platforms), content production, agency retainers, and sometimes a slice of product-marketing headcount. It's what your CFO uses and what acquirers diligence. Fully-loaded CAC at a Series B SaaS company is typically 1.5x to 2.5x the paid CAC, depending on team size.
Load-bearing trick: When someone quotes a CAC number, always ask which one. A founder pitching "$80 CAC" when paid-only is $80 but fully-loaded is $240 isn't lying — they're just using the most flattering definition. The board cares about fully-loaded; the channel manager cares about paid.
LTV/CAC and why the magic number is 3
The single ratio every growth lead memorizes: LTV/CAC ≥ 3. The intuition is rough but durable. One unit of CAC is the spend itself. One unit covers the gross-margin cost of serving the customer (hosting, support, payment processing). The third unit is the actual contribution to fixed costs and profit. Anything below 3 means you're funding growth out of pocket; anything above 5 usually means you're underspending and a competitor will outbid you for the same channels.
| LTV/CAC ratio | Interpretation | Typical action |
|---|---|---|
| < 1 | Every customer is a net loss | Stop spending, fix pricing or retention |
| 1 – 2 | Underwater on payback | Cut weakest channels, raise prices |
| 3 | Healthy floor | Maintain, optimize mix |
| 3 – 5 | Strong unit economics | Press accelerator, expand channels |
| 5+ | Probably underinvesting | Test new channels, raise ad budgets |
One caveat the ratio hides: LTV must be margin-adjusted, not revenue-adjusted. A consumer marketplace like DoorDash with 20% contribution margin and revenue LTV of $400 has a true LTV of $80 — so a $30 CAC is fine but a $90 CAC is catastrophic. Half the LTV/CAC debates on Twitter come from people quoting revenue LTV against fully-loaded CAC. Different denominators, different planet.
CAC payback by industry
CAC payback period answers a different question than LTV/CAC: how many months until the customer pays back what we spent to acquire them? It's a cash-flow question, not a profitability question, and it matters more when capital is expensive (2023–2026) than when money is free (2020–2021).
Sanity check: CAC payback = CAC ÷ (monthly ARPU × gross margin). A $600 CAC on a $50/month plan at 80% margin pays back in 15 months. The same CAC on a $50/month plan at 20% margin (think hardware-bundled SaaS) pays back in 60 months — which is a death sentence for a Series A.
Reasonable payback benchmarks vary wildly by business model. Consumer subscription apps (Headspace, Calm, Duolingo) target 3–6 months because churn is fast and capital is expensive. SMB SaaS (Notion teams plan, Linear Standard) targets 6–12 months. Mid-market SaaS (Snowflake mid-segment, Databricks) tolerates 12–18 months because annual contracts and NDR 120%+ make the math work. Enterprise SaaS (Workday, ServiceNow) accepts 18–24 months or longer because a single Fortune 500 logo compounds for a decade.
When investors talk about "the rule of 24" they mean: payback over 24 months is a yellow flag at any stage, and a red flag at growth-stage. That's the bar to internalize.
CAC by channel — realistic ranges
Channel-level CAC is where strategy gets concrete. Aggregate CAC is a vanity metric — channel CAC is the operating metric. The ranges below are typical US B2C-to-mid-market SaaS in 2026; your numbers will differ but the ordering tends to hold.
| Channel | Typical CAC range (US, 2026) | Quality / intent | Notes |
|---|---|---|---|
| Paid search (Google) | $80 – $400 | High intent | Scales with bid; CPCs creeping up post-AI search |
| Paid social (Meta, TikTok, LinkedIn) | $60 – $350 | Mid intent | Creative-fatigue cycle ~2 weeks; LinkedIn 3-5x higher |
| SEO / content | $20 – $120 | High intent | Long ramp (6–18 months); compounds after critical mass |
| Referrals / word-of-mouth | $10 – $80 | Highest intent | Cheapest but caps out; needs a real loop to scale |
| Outbound / SDR (B2B) | $400 – $2,500 | Mid intent | Only sane path for enterprise ACV > $30k |
A few patterns to internalize. Referrals dominate on CAC but cap on volume — Dropbox famously scaled to 100M users on a referral loop but every company that copied the playbook discovered the loop needs an underlying product worth referring. SEO is the slowest and the cheapest at scale, which is why content moats matter when ZIRP ends. Outbound only pencils out at high ACV — a $400 outbound CAC on a $1,200/year SMB SaaS deal is a slow death; the same CAC on a $50k enterprise contract is a no-brainer.
The blended CAC of a healthy mid-stage SaaS is usually a weighted average around $150 – $400, with referrals and SEO pulling the average down and outbound pulling it up.
Common pitfalls
The pitfall that sinks more boards than any other is ignoring fully-loaded CAC. A growth team will report "$120 paid CAC" with a straight face while quietly omitting the $80k/month growth team payroll, the $40k/year HubSpot bill, and the agency retainer. When fully-loaded numbers finally land in a board meeting, the real CAC is 2x and the LTV/CAC ratio quietly fails. The fix is to publish all three flavors — blended, paid, fully-loaded — on the same dashboard and force the team to discuss the gap.
A close second is attribution naivety. Most teams still default to last-click, which over-credits whatever channel happens to be at the bottom of the funnel — usually paid search or retargeting. A typical SaaS buyer touches the brand 5–9 times across blog posts, podcasts, peer mentions, and a final search before converting. Last-click attribution would zero out the blog and credit Google entirely. The fix is to run multi-touch attribution or at minimum compare last-click against linear and time-decay models — if the three disagree by more than 30%, you're making channel-budget decisions on noise.
A subtler trap is comparing revenue LTV to fully-loaded CAC without a margin adjustment. A founder will say "LTV is $1,800 and CAC is $400, so we're at 4.5x." But if gross margin is 35% — common in marketplaces or hardware-bundled SaaS — the margin-adjusted LTV is $630, and the real ratio is 1.6x. That's a business that looks healthy on a pitch deck and is actually burning. Always apply gross margin (or contribution margin if you can) before computing the ratio.
A fourth pitfall: picking too short an LTV window. Historical 30-day or 90-day LTV is fine for fast consumer apps but useless for B2B SaaS where the median customer life is 36+ months. Using 90-day LTV against a 12-month payback target will make every channel look unprofitable. This is why mature growth teams use predicted LTV from a cohort survival model, not trailing-window LTV.
Finally, forgetting to blend organic into your reported CAC flatters paid channels at the company's expense. If 40% of your new customers come from SEO and word-of-mouth, your "paid CAC" can be $300 while your blended CAC is $180 — and the business is healthier than the paid number suggests. Conversely, when organic share drops because Google's AI Overviews ate your top funnel, blended CAC will spike before paid CAC moves. Watching the gap between paid and blended is one of the highest-signal habits in growth.
Related reading
- How to calculate CAC in SQL
- How to calculate blended CAC in SQL
- How to calculate CAC by channel in SQL
- How to calculate CAC payback in SQL
- How to calculate LTV in SQL
- Growth loops primer for PMs
If you want to drill metric-design and SQL questions like this every day, NAILDD is launching with 500+ analytics problems across exactly these patterns.
FAQ
What's the difference between CAC and CPA?
CPA (cost per acquisition) is usually channel-specific and refers to any conversion event — a signup, a lead, a trial start. CAC is reserved for a paying customer and is typically reported at the company level. The shorthand: CPA lives in the ad platform UI, CAC lives in the board deck. When a media buyer says "$30 CPA on Meta," they probably mean cost per trial signup, not cost per paying customer.
Should CAC include free-trial users or only paid conversions?
For SaaS with a free trial, CAC should be measured against paying customers, not trial signups — otherwise you'll under-report and confuse the trial-to-paid conversion problem with the acquisition problem. The exception is freemium products like Notion or Slack where free users are a load-bearing part of the growth loop; in that case, run two CACs: one against activated free users (for the loop), one against paid (for the P&L).
How do I handle organic customers in CAC?
For blended CAC you include them in the denominator (they make the number look better and that's appropriate — they're real customers). For paid CAC you exclude them. For fully-loaded CAC you include them but you also include the cost of producing the organic acquisition machine — content writers, SEO tools, PR retainers. Pure organic CAC is never zero; it just sits on a different line in the budget.
What attribution window should I use for CAC?
Use the window that matches your sales cycle. Consumer apps with same-day conversions: 7 days. SMB SaaS with a 2-week trial: 30 days. Mid-market with a 60-day evaluation: 90 days. Enterprise with a 6-month procurement cycle: 180 days or full multi-touch. The mistake is using the platform default (typically 7-day click on Meta, 30-day click on Google) when your real cycle is months long — you'll wildly under-credit upper-funnel channels.
How is CAC payback different from LTV/CAC?
LTV/CAC is a profitability ratio — does this customer make us money over their lifetime? CAC payback is a cash flow metric — how many months until we get our acquisition spend back? You can have great LTV/CAC and terrible payback (enterprise deals with annual contracts), or great payback and mediocre LTV/CAC (consumer apps with high churn after month 6). Both matter, and they answer different boardroom questions.
Can CAC be negative?
In rare cases, yes — usually for products with strong referral mechanics where acquired customers immediately bring in 1.5+ new customers through a viral loop. Early Dropbox, early Hotmail, and some embedded-payments products had effectively negative CAC during their hyper-growth phases. It doesn't last; viral coefficients regress to the mean once the obvious adopters are saturated.