The CAC formula marketers keep getting wrong

A SaaS company reports $500 customer acquisition costcustomer acquisition costCustomer Acquisition Cost (CAC) is the total sales and marketing spend divided by the number of new customers gained in a period. It measures how efficiently you grow.View full definition → (CAC) on its board deck. Investors nod. Six months later, finance recalculates with fully loaded costs and cohort timing, and the real number is $1,400. Nobody lied. They used the formula everyone uses, and that formula flatters almost every company that runs it.

CAC is the object the rest of this module sits on: efficiency ratios, channel budgets and pricing all inherit whatever errors are baked into it. In software and SaaS (Software as a Service, meaning software sold via subscription rather than one time licence), an understated CAC means overfunding a dead channel, underpricing the product, or watching a diligence team arrive at a number three times yours.

The naive formula, and why it lies

The version most people learn first:

CAC = Total Sales & Marketing spend ÷ New customers acquired

Simple. Also wrong most of the time, for three reasons.

1. Timing mismatch. You spend this month. The customers that spend produces sign next month, next quarter, or after a nine month enterprise cycle. Dividing this month's spend by this month's new logos compares two unrelated populations.

2. Cohort blending. A company running $200 CAC on self-serve and $8,000 on enterprise deals might report a blended $600, a number that describes neither business. Segment ran both at once for years: a free developer tier alongside a sales team calling on data engineering leaders. One average across those two motions would have been arithmetic, not information.

3. Spend scope. Paid mediaPaid mediaVisitors arriving via paid ads or sponsored placements, where you pay a platform to display your message rather than earning visits organically.View full definition → only? Or sales salaries, commissions, the marketing team, the CRMCRMCustomer Relationship Management: software and strategy to manage and analyse customer interactions throughout their lifecycle.View full definition → and analytics stack, content production, events, the booth? Sector benchmarks assume fully loaded S&M (sales and marketing) spend. Plenty of companies quietly report media only, and the gap is usually a multiple, not a rounding error.

The formula that survives scrutiny

Fully loaded CAC (cohort-aligned) = (Total S&M spend in period T) ÷ (New customers who signed contracts attributable to period T's spend)

The fix: align spend to the cohort it produced, not the cohort that happened to close in the same calendar month.

Worked example

A mid-market SaaS company spends $300,000 on sales and marketing in Q1 (salaries, ads, tools, events, all in). Average sales cycle is 90 days, so Q1's leads convert mostly in Q2.

Naive calculation: Q1 closed 40 deals, mostly from Q4's pipelinepipelineAll active sales opportunities across the stages of the sales process, together with their combined potential value and probability of closing.View full definition →, so $300,000 ÷ 40 = $7,500 CAC. Looks fine.

Cohort-correct calculation: Q1's spend produced 25 customers who signed in Q2. CAC = $300,000 ÷ 25 = $12,000.

That is 60% higher, and it is the real cost of that cohort. Pricing built on $7,500 is pricing built on a cohort that does not exist.

What belongs in the numerator

Fully loaded means the accounting S&M line plus several costs accounting deliberately puts elsewhere.

• Payroll loading. Employer taxes, benefits and equity push the real cost of a rep well above base salary, commonly by a fifth to a third. CAC models built from base salaries understate every sales-led channel.
• Capitalised commissions. Under ASC 340-40, incremental costs of obtaining a contract are capitalised and amortised over the expected customer life, so the reported S&M expense lags the cash you actually paid to win this year's logos. Board CAC should use cash commissions on new business; the two diverge most in the years you are growing fastest, which is exactly when the number matters.
• Partner and marketplace economics. Klaviyo, which sells email and SMS marketing software, acquires a large share of its customers through the Shopify app marketplace and agency partners. Whatever a company pays for that access, revenue sharerevenue shareThe percentage of total industry sales your company captures in a given period. It measures competitive position relative to rivals in a defined market.View full definition →, referral fees, partner enablement, is acquisition cost even when it lands in cost of revenue.
• First-year discounts. A 30% discount to close a competitive deal is acquisition spend with a different label. It never touches the S&M line, and it makes the CAC of a heavily discounted enterprise cohort look identical to a full-price one.

The denominator rots just as easily. Counting free signups or trials as customers inflates it. Counting expansions and seat upgrades as new logos inflates it again, since expansion is retention work, not acquisition. And counting inboundinboundA strategy that attracts prospects organically via valuable content (blog, SEO, social) rather than interrupting them.View full definition → word-of-mouth signups against paid spend produces a blended CAC that looks efficient right up to the moment you try to buy more of it. Report paid CAC and blended CAC as separate lines, or the marginal cost of the next customer stays invisible.

Break it out before you average

Blended CAC is a starting point, not a decision. Break it out by:

• Channel: paid search, organic and SEOSEOSearch Engine Optimization: the practice of improving your pages' natural (unpaid) rankings in search engine results pages to attract more organic traffic.View full definition →, outboundoutboundProactive outreach that pushes your message to targeted audiences through advertising, email, or direct prospecting, initiated by the seller rather than the buyer.View full definition →, partnerships, self-serve
• Segment: SMB (small and medium business), mid-market, enterprise
• Region: US CAC often runs higher than European CAC for comparable segments, largely because of ad auction prices and sales compensation norms (a directional pattern, not a fixed ratio)

Monday.com spent roughly as much on sales and marketing as it booked in revenue around 2021. That is defensible only if the number decomposes into channels with visibly different economics, so the board can see which line to cut and which to feed. A single blended figure at that spend level tells a board nothing except that the money left the building.

None of this is yet a verdict on whether the spend was smart. That happens when you divide CAC by monthly gross profit per customer and read the result against the thresholds the benchmarking lesson sets. The job here is making sure the number entering that division is honest, because a CAC understated by 40% turns a genuinely poor 20 month payback into a comfortable-looking 12.

LTV:CAC, and why the "3:1 rule" is oversimplified

LTV:CAC ratio = Customer Lifetime Value ÷ CAC

The modelling on the numerator, cohort curves, margin treatment and horizon caps, has its own lesson. The point here is that the ratio inherits every allocation error below it, and inherits them multiplicatively. Understate CAC by 40% and a 2:1 ratio presents as 3.4:1, which clears the commonly cited threshold without a single thing having changed in the business.

Treat 3:1 as a rough heuristic (cited by VCs and reports like OpenView's SaaS Benchmarks, estimates, figures vary by year). Below 1:1 you lose money on every customer, and that floor is the reliable half of the rule. A company with 130% net revenue retention (NRR, meaning existing customers expand faster than they churn) can live with a lower ratio because the cohort keeps growing after acquisition.

A minimal script to align cohorts correctly

For teams pulling this from a CRM (customer relationship management system) or data warehouse, the core logic is a cohort join, not a same-period average:

-- Simplified cohort CAC logic
SELECT
  spend_month,
  sm_spend,
  signed_customers_from_this_spend,
  sm_spend / signed_customers_from_this_spend AS cohort_cac
FROM (
  SELECT
    s.month AS spend_month,
    s.total_sm_spend AS sm_spend,
    COUNT(c.customer_id) AS signed_customers_from_this_spend
  FROM sm_spend_by_month s
  JOIN customers c
    ON c.attributed_spend_month = s.month  -- attribution model, not signup month
  GROUP BY s.month, s.total_sm_spend
) t;

The critical column is attributed_spend_month, tying each customer back to the pipeline period that generated them. Segment, which sells the customer data infrastructure many teams use to assemble exactly this join, is a reminder that the attribution model is a modelling choice someone has to own, not a fact the warehouse hands you.

What to actually flag in board reporting

• Report CAC by segment and channel, never a single blended figure without at least a footnote
• State whether the number is fully loaded (salaries, tools, overhead, partner fees) or media-only
• State whether commissions are cash paid or the amortised expense, since the two diverge in growth years
• Note the attribution window used (30, 60, 90 days), since B2B (business to business) deals rarely close in the month the spend occurred

🎬 [VIDEO: "SaaS Metrics 101: CAC, LTV, and Payback Explained" - youtube.com/results?search_query=saas+metrics+cac+ltv+payback - a practical walkthrough of how these formulas connect for founders and marketers building their first metrics dashboard]

Key Takeaways

• Naive CAC (spend ÷ new customers, same period) understates true cost by ignoring the lag between spend and signed contract.
• Fully loaded means loaded payroll, cash commissions, partner and marketplace fees and first-year discounts, not just media, and aligned to the cohort the spend produced.
• The denominator lies too: free signups, expansions and word-of-mouth logos counted against paid spend all deflate CAC. Keep paid CAC and blended CAC on separate lines.
• Segment by channel and customer tier before averaging; a blended figure at high spend levels tells a board nothing actionable.
• Every downstream ratio inherits these errors and magnifies them, so a 40% understatement is enough to manufacture a healthy-looking business out of an unhealthy one.