How to Analyze Customer Acquisition Cost and Lifetime Value for Growth Stocks

CAC and LTV: The Foundation of Growth Company Valuation

Most investors value growth companies on revenue multiples. Revenue is easy to track, easy to forecast, and it shows up in every headline. The problem is that revenue tells you nothing about whether growth is profitable at the unit level. A company growing revenue at 50% annually while spending $2 to acquire every $1 of customer value is not building a business — it is running a sophisticated wealth transfer from shareholders to customers.

This is precisely what happened in the 2019–2022 ZIRP bubble. Venture-backed companies weaponized cheap capital to subsidize customer acquisition, and public market investors rewarded them with premium EV/Revenue multiples because the top line looked spectacular. Peloton spent $1,100 in fully loaded acquisition cost per subscriber generating $39/month in gross profit — a 28-month payback on a product with average retention of 18 months. The unit economics were underwater from day one. DoorDash spent $35–$40 to acquire each consumer order, which generated approximately $5 in gross profit, requiring 7–8 orders before payback — and average order frequency was declining. These were not viable businesses at those economics, regardless of revenue growth.

By contrast, the companies that created lasting shareholder value through that same period — CrowdStrike, Shopify, HubSpot, Datadog — all showed a consistent pattern: improving CAC efficiency, expanding LTV through product cross-sell, and payback periods under 18 months. These were not just growing. They were growing profitably at the unit level, and the unit economics predicted the stock performance years in advance.

How to Calculate CAC: Blended, Fully Loaded, and Channel-Level

Customer acquisition cost seems straightforward. Divide total sales and marketing expense by new customers acquired. In practice, there are three versions, and using the wrong one can lead to catastrophically bad investment decisions.

Blended CAC

Total S&M expense divided by total new customers. This is the most commonly cited version and also the most misleading. It includes organic and word-of-mouth customers in the denominator (who cost essentially nothing to acquire), which subsidizes the CAC of paid customers. A company might report blended CAC of $200 when its paid acquisition CAC is $600 because 65% of new customers come through organic channels. The investment risk is that organic acquisition often plateaus as the company scales, forcing heavier reliance on paid channels with much worse economics.

Fully Loaded CAC

This adds stock-based compensation allocated to sales and marketing, SDR and BDR fully burdened cost, sales engineering support, and allocated overhead. For high-growth software companies, SBC alone can represent 20–35% of total S&M spend. CrowdStrike's FY2025 S&M expense was approximately $1.05 billion on a GAAP basis (including ~$280 million in SBC). Excluding SBC as many companies encourage analysts to do would understate the true cost of acquiring customers by 27%. If you are not adjusting for SBC-loaded CAC, your LTV/CAC ratio is fiction.

Channel-Level CAC

The most granular and most useful version. This breaks down acquisition cost by channel: paid search, paid social, content marketing, direct sales team, channel partners, and organic. Most public companies do not disclose this level of detail, but you can infer it from earnings call commentary, investor day presentations, and the trajectory of S&M spend relative to customer adds. When management says “we are seeing improving efficiency in our go-to-market motion,” press them on which channels are improving. If all the improvement is coming from organic growth flattening the blended number while paid CAC is rising, the trend is unsustainable.

LTV Calculation Methods: Historical Cohort vs. Predictive

Lifetime value is where most analysts go wrong, not because the math is hard but because the assumptions are hidden and often heroic.

The Predictive Formula (and Its Traps)

The standard predictive LTV formula is: ARPA × Gross Margin × (1 / Annual Churn Rate). This formula assumes stable retention and pricing in perpetuity. If a SaaS company has $25,000 ARPA, 82% gross margin, and 8% annual churn, predictive LTV is $25,000 × 0.82 × (1/0.08) = $256,250. That looks spectacular. But consider: an 8% annual churn rate means the median customer lifetime is 12.5 years. Is it realistic that a customer will stay for 12.5 years? For Salesforce or Oracle, perhaps. For a Series B startup competing in a crowded market, probably not.

The more sophisticated approach applies a discount rate and caps the lifetime at 5–7 years: LTV = ∑ (ARPA × Gross Margin × Retention Rate^n) / (1 + Discount Rate)^n, summed from n=1 to n=5. This accounts for the time value of money and the reality that distant-future retention rates are inherently uncertain. With a 10% discount rate and 5-year cap, the same $256,250 predictive LTV drops to roughly $74,000. That is a 71% haircut — and in our experience, the capped number is closer to economic reality.

Historical Cohort Analysis (The Gold Standard)

Cohort analysis tracks the actual behavior of customer groups over time. HubSpot, to their credit, discloses cohort economics in their annual investor presentations. Their 2018 cohort generated over $6 of cumulative gross profit for every $1 of acquisition cost by year five. Each successive vintage showed faster payback and higher cumulative value, confirming the flywheel. This is exactly what you want to see: older cohorts validating the model, and newer cohorts improving on it.

When companies stop disclosing cohort data — or never start — that is a red flag. It usually means the cohorts do not tell a good story. If management cannot show you that customers acquired three years ago are generating more revenue today than when they were acquired, question whether the LTV projections embedded in the valuation are realistic.

Payback Period: The Real Constraint on Growth

Here is a truth the growth-at-all-costs crowd ignores: LTV/CAC ratio tells you whether growth is profitable. Payback period tells you whether the company can afford to grow. A 6x LTV/CAC with a 30-month payback requires the company to finance 2.5 years of customer acquisition upfront before recovering its investment. At scale, this means every $100 million in annual S&M spend creates a $250 million working capital deficit that must be funded with cash on hand, debt, or equity dilution.

Payback period is calculated as: CAC / (ARPA × Gross Margin / 12). If fully loaded CAC is $30,000, ARPA is $24,000, and gross margin is 80%, monthly gross profit per customer is $1,600, and payback is 18.75 months. The benchmarks vary by business model:

CAC/LTV Predicted the Winners and the Losers

The beauty of unit economics is that they tell you the truth about a business before the income statement does. Revenue can grow while unit economics deteriorate, and during those periods the stock can rally on headline growth. But eventually the unit economics win. Always.

Shopify: The Textbook Case

In 2017, Shopify's S&M expense was approximately $310 million, supporting net new merchant adds of roughly 200,000. Blended CAC: approximately $1,550. But Shopify's real magic was on the LTV side. Merchants who started with a $29/month Basic plan expanded into Shopify Plus ($2,000+/month), adopted Shopify Payments (which added 2.4–2.9% of GMV in revenue), used Shopify Capital, and integrated Shopify Fulfillment. A merchant acquired in 2017 at $1,550 CAC was generating $8,000–$12,000 in annual revenue by 2021, with net revenue retention above 110%. The LTV/CAC ratio improved from approximately 4x in 2017 to 7x+ by 2020. The stock went from $90 to $1,760 peak. The unit economics were visible to anyone who calculated them.

WeWork: The Anti-Pattern

WeWork's S-1 filing in 2019 revealed a business spending an estimated $7,000–$10,000 per desk in acquisition and build-out costs. The average desk generated approximately $6,500 in annual revenue at 20–25% contribution margin — roughly $1,300–$1,625 in annual gross profit per desk. Payback: 4.3–7.7 years. But WeWork's average lease term was 15 years while the average member stayed for 15 months. The mismatch was fatal. The company was acquiring customers with 4+ year payback periods on a product with 15-month average retention. LTV/CAC, properly calculated, was below 1x. Every desk WeWork filled destroyed value. Revenue grew from $886 million in 2017 to $3.2 billion in 2019, but the unit economics screamed insolvency the entire time.

The SBC Adjustment That Changes Everything

Stock-based compensation is a real cost. Full stop. When a company issues shares to sales reps as part of their compensation, those shares dilute existing shareholders just as surely as cash compensation would reduce earnings. Yet the tech industry has collectively agreed to pretend SBC does not exist when calculating “adjusted” metrics, and sell-side analysts have largely complied.

Consider this worked example. A mid-cap SaaS company reports $500 million in total S&M expense (GAAP), of which $120 million is SBC. They acquired 5,000 net new customers. The company's investor presentation shows CAC of $76,000 using non-GAAP S&M ($380M / 5,000). But fully loaded CAC is $100,000 ($500M / 5,000) — 32% higher. If ARPA is $60,000 and gross margin is 78%, monthly gross profit is $3,900 and non-GAAP payback is 19.5 months. But GAAP-loaded payback is 25.6 months. At an 8% annual churn rate and 5-year LTV cap, the non-GAAP LTV/CAC ratio is 4.1x. The GAAP-loaded version is 3.1x. The non-GAAP version looks like a well-oiled SaaS machine. The GAAP version looks like a company whose unit economics barely clear the 3x threshold. The investment decision might be different.

The companies with the healthiest unit economics — Veeva Systems, MSCI, S&P Global — have SBC as a low single-digit percentage of revenue. When you see SBC above 15% of revenue, the gap between reported and real unit economics is large enough to flip the investment thesis. Always check. For a deeper framework on evaluating software company metrics specifically, see our guide on SaaS metrics for software stock analysis.

Why Improving Unit Economics Matters More Than Revenue Growth

This is the contrarian take that the market consistently underweights: a company growing revenue at 25% with improving unit economics will outperform a company growing revenue at 40% with deteriorating unit economics over any 3+ year horizon. The reason is mathematical. Revenue growth without unit economics improvement is a treadmill — the company must run faster and faster just to maintain its cash burn rate. Revenue growth with improving unit economics is a flywheel — each new customer is cheaper to acquire and more valuable over time, which funds further growth, which further improves economics through scale.

Datadog illustrates this perfectly. Between 2020 and 2025, Datadog's revenue growth decelerated from 66% to approximately 25%. The stock underperformed growth benchmarks during the deceleration. But fully loaded CAC declined roughly 20% as the sales force matured and product-led growth gained traction, while LTV expanded through module adoption (the average enterprise customer uses 4+ products, up from 2 in 2020). The LTV/CAC ratio improved from approximately 4x to 6.5x. Net revenue retention stayed above 120%. The unit economics told you the business was getting stronger even as the growth rate slowed. The stock reflected this by re-rating from 15x NTM revenue at the trough to 22x as the market recognized the margin expansion potential embedded in improving unit economics.

The practical lesson: when screening growth stocks, filter first for improving unit economics, then for revenue growth. A company that can demonstrate 4+ quarters of declining CAC payback period and expanding LTV/CAC ratio is building durable competitive advantage. A company that can only demonstrate accelerating revenue may just be spending more on sales and marketing.

Frequently Asked Questions