MarketingGrowth & Acquisition

Getting CAC and LTV honest before scaling

Most growth failures trace back to a single mistake: scaling on metrics that look right but are calculated wrong. This article breaks down how to build CAC and LTV figures you can actually trust before committing budget.

🎙️

Listen to the podcast

4 min

Customer acquisition cost and lifetime value are the two numbers that determine whether a business scales into profit or scales into a faster version of losing money. CMOs know both terms. Fewer have clean, defensible definitions of either sitting inside their organisation right now. That gap is where scaling decisions go wrong.

The confusion is not about ignorance. It comes from the fact that CAC and LTV are ratios, not facts. They depend entirely on what you choose to include in the numerator, what time horizon you use, what assumptions you make about churn and margin. Two companies in the same sector can report wildly different figures using perfectly legitimate accounting choices. The problem arrives when those choices are optimistic rather than honest, and someone uses the output to justify a tenfold increase in paid media spend.

Why it matters specifically for CMOs

A CFO or board member evaluating a growth case will eventually stress-test your numbers. If your CAC excludes your customer success team's onboarding costs, or your LTV is calculated on gross revenue rather than gross margin, the model collapses under scrutiny. At that point, the conversation stops being about strategy and becomes about credibility.

There is a more immediate risk. Scaling a paid acquisition channel while working from an inflated LTV-to-CAC ratio means every new customer you acquire is potentially worth less than you paid to get them. The cash flow damage compounds before the measurement error becomes visible. By the time cohort data catches up, you have already allocated the next quarter's budget on the same broken assumption.

Casper, the mattress company, is an instructive case. Through its high-growth phase in the late 2010s, the business spent aggressively on digital acquisition with LTV assumptions that proved optimistic once actual repeat purchase rates and return costs were factored in. The unit economics that looked viable in the model did not survive contact with real customer behaviour at scale. The company went public in 2020 at a fraction of its private valuation and was taken private again two years later. The acquisition math mattered.

How it actually works: the mechanics

CAC is total sales and marketing spend divided by the number of new customers acquired in the same period. The fight is always over what goes into "total sales and marketing spend." A conservative, honest version includes salaries and benefits for everyone in sales, marketing, and customer success who touches the pre-conversion journey, agency and contractor fees, tools and software, creative production, and any channel spend. A flattering version includes only the media spend. The difference can easily be a factor of two or three.

Take a concrete example. A SaaS company spends 200,000 dollars on paid search in a quarter and acquires 400 customers. Headline CAC: 500 dollars. Add 80,000 dollars in sales team salaries, 30,000 in tooling, and 20,000 in creative: total spend is 330,000 dollars, real CAC is 825 dollars. If the sales team uses a 500-dollar CAC figure when pitching a budget increase, they are pitching a business that does not exist.

LTV is the net present value of all future margin a customer will generate over their relationship with the business. The most common mistake is using revenue where you should use gross margin. A customer generating 1,200 dollars per year at 30% gross margin has a revenue LTV of 1,200 dollars times their expected lifespan. Their margin LTV is 360 dollars times that lifespan. If your CAC is 700 dollars and you are comparing it to a revenue-based LTV, the ratio looks fine. Compared to the margin LTV, you may never recover the acquisition cost.

The discount rate matters too, especially in a higher interest rate environment. Money recovered over five years is worth less than money recovered in eighteen months. Ignoring the time value of cash in LTV calculations overstates what a customer is actually worth to the business today.

A workable formula: LTV = (average annual gross margin per customer) divided by (annual churn rate), with a discount factor applied if the relationship is expected to be long. For a B2B SaaS company with 80-dollar average monthly gross margin per account, 15% annual churn, and a 10% discount rate, the LTV is roughly 480 dollars, not the 960 dollars you get if you simply multiply monthly revenue by an assumed 12-month lifetime.

When to use it and when not to: the honest tradeoffs

CAC and LTV are most useful when you have at least 12 months of cohort data, stable product economics, and a clearly defined customer segment. At that point, you can measure churn, you understand margin by channel and cohort, and the model is grounded in observation rather than projection.

They are least useful, and most dangerous, in three situations. First, in the first 12 to 18 months of a new product or market entry, before you have real retention data. Whatever LTV you calculate then is a hypothesis. Treat it as one. Second, when your customer base is heterogeneous and you are averaging across very different segments. Enterprise customers and SMB customers in the same LTV calculation will produce a number that accurately describes neither. Third, when your cost structure is changing rapidly, because historical CAC figures stop predicting future ones the moment you shift channel mix, hire aggressively, or change your pricing model.

The honest tradeoff is that detailed, accurate CAC and LTV calculations take time and require finance and data collaboration that many marketing teams do not have. A simpler, rougher calculation done on accurate inputs is more useful than a sophisticated model built on assumptions chosen for their optics. Airbnb's growth team was known for running tight unit economics reviews tied to actual margin data rather than revenue proxies. That discipline is replicable regardless of company size.

The single most valuable thing a CMO can do before presenting a scaling case is to have someone who was not involved in building the model try to break it. If the model does not survive that review, it will not survive the scale. Fix it before you spend the money.

Finished reading?

Validate your read to earn XP and feed your radar.