What are CAC, LTV, PBP in marketing?

CAC, LTV, and Payback Period are the three metrics that decide whether a growth motion is profitable or structurally subsidized. They're simple to define and easy to misuse — most of the "our LTV:CAC is 5:1" claims you'll see in pitches don't hold up under scrutiny.

The three metrics

• CAC — Customer Acquisition Cost. Total spend to acquire a customer. Includes paid ads, sales team salary + commissions, content marketing amortized, onboarding costs attributable to acquisition. The purist definition includes everything spent to get the customer to first value; the lazy definition is just ad spend.
• LTV — Lifetime Value. Estimated net profit per customer over the full relationship. LTV = (ARPU × Gross Margin) / Churn Rate for a simple subscription model. Note it's gross profit, not revenue — a common source of overstated LTV.
• PBP — Payback Period. Months it takes for gross profit from a customer to cover CAC. PBP = CAC / (ARPU × Gross Margin / 12) for monthly subscriptions.

The LTV:CAC ratio

A rough readout on business-model health:

LTV:CAC	Interpretation
1:1	Losing money. The more you sell, the more you lose.
2:1	Marginal. Works if your CAC is very short payback and churn is low.
3:1	Healthy. Most successful SaaS businesses land here.
5:1+	Under-investing in marketing. You could grow faster by spending more.
10:1+	Either misreporting or leaving enormous growth on the table.

The 3:1 target is surprisingly durable — it shows up in Bessemer's cloud index data, in public SaaS companies, and in most consumer-subscription models. The logic is that by the time you net out overhead, R&D, and G&A (which aren't in the CAC calculation), a 3:1 ratio leaves enough margin to run a real company.

The common manipulations

Most "great unit economics" numbers fall apart if you check three things:

1. How CAC is calculated

The honest CAC includes every dollar spent to bring a customer to first paid usage:

CAC = (Paid Media + Sales Salaries + Sales Commissions +
       Marketing Team Salaries + Content/Brand Spend +
       Free Trial Costs + Onboarding Costs) / New Customers Acquired

The common sloppy version is Ad Spend / New Customers, which excludes 60-80% of the real cost for most B2B SaaS companies. If someone reports CAC = $500 for an enterprise sale, ask what's in the numerator.

2. How LTV is calculated

Three frequent errors that inflate LTV:

• Using revenue instead of gross profit. If you sell a $100/mo subscription with 70$70/mo, not $100/mo`. Using revenue inflates LTV by ~30%.
• Using an optimistic churn rate. Many teams use the current month's churn (which is often lower than steady-state) rather than cohort-based churn. Early-cohort churn is a poor predictor for SaaS products where the cancel button is easy to reach in month 2-3.
• Ignoring cohort decay. 1/churn assumes exponential decay forever. Real retention curves often flatten after a year, which means LTV is actually higher than the simple formula — but it can also be lower if late-stage churn spikes (e.g. contract-renewal cliff).

3. Whether LTV is realized or projected

An LTV of $5,000 based on a 2-year retention cohort is a claim. An LTV of$5,000 based on 3 months of data and an assumed churn rate is a guess. Most startup LTV numbers are guesses. Investors and disciplined operators discount them by 30-50% until 12+ months of retention data is available.

Why payback period is the metric that actually matters

In practice, I care about payback period more than LTV:CAC for early and mid-stage companies. Three reasons:

1. Payback doesn't require a churn estimate. You can measure payback with 6 months of data. You can't measure LTV with 6 months of data — you have to project churn, which is the part most frequently wrong.
2. Payback maps directly to cash runway. A 6-month payback means a dollar spent on CAC comes back in 6 months and can be redeployed. A 24-month payback means every growth dollar is effectively an investment that ties up working capital for 2 years. Same LTV:CAC, very different cash dynamics.
3. Payback is falsifiable quickly. LTV projections take years to falsify. Payback claims can be checked against the P&L in a quarter.

Rule of thumb for SaaS:

• < 12 months payback: Excellent, can grow aggressively
• 12-18 months payback: Healthy, typical for enterprise SaaS
• 18-24 months payback: Requires cheap capital and patient investors
• > 24 months payback: Structurally subsidized growth; shouldn't be scaled without a plan to reduce it

Segment-level unit economics

A single company-wide LTV:CAC ratio is usually an average of wildly different segments. The more useful analysis:

• Break out by channel. Paid search CAC vs content-marketing CAC vs outbound-sales CAC are often 5-10x different. A 3:1 blended ratio might hide 8:1 on content and 0.8:1 on paid search.
• Break out by customer size. Enterprise, mid-market, and SMB usually have different CAC, different churn, and different expansion. A blended LTV:CAC can mask a broken SMB motion.
• Break out by cohort. Newer cohorts often have lower CAC (brand is working) or higher CAC (you've exhausted cheap channels). Either trend is worth catching early.

If you're reporting unit economics to a board or investor and the number is a single blended ratio, the next question is always "by segment?" Expect it, be ready for it.

The expansion multiplier

For SaaS with net revenue retention > 100%, LTV is meaningfully higher than ARPU × GM / churn because existing customers expand faster than they churn. A company with 120% NRR has negative net churn, and LTV becomes mathematically unbounded in the simple formula. Real calculations cap it at a reasonable horizon (e.g. 7 years) or use a DCF.

Companies with strong expansion dynamics (Snowflake, Datadog, MongoDB) can support payback periods that would look alarming for flat-NRR companies, because every month the cohort grows in value rather than decaying.

Practical dashboard

A minimally honest unit-economics dashboard should show:

1. CAC by channel, trailing 6 months.
2. Payback period by channel and by segment.
3. Cohort retention curves (gross and net) by acquisition month.
4. LTV:CAC by segment, clearly labeled as "projected" if based on <12 months of data.
5. Net revenue retention trailing 12 months.

Most companies ship only (4), and often only the blended version. The others are where the real signal lives.

See also

• What is a Market? — CAC varies radically by market/segment; a blended number hides the good and bad reference networks.
• MMRs, Neutralizers and Differentiators — the features driving retention (and therefore LTV) are almost always differentiators, not MMRs.
• 9x Effect — a product that clears the 9x threshold typically has lower CAC and higher LTV because references compound faster.