Core Marketing Metrics: CAC, ROAS, LTV, and Conversion Rate

1) Customer Acquisition Cost (CAC)

What it measures

CAC estimates how much you spend to acquire one new customer in a defined period and scope. It is most useful when paired with payback thinking: not just “is CAC low?” but “how quickly do we earn it back?”

Formula

Media-only CAC (sometimes called “blended media CAC” if it includes all paid channels):

CAC_media = Paid media spend / New customers acquired from paid media

Fully-loaded CAC (captures the true cost to acquire):

CAC_fully_loaded = (Paid media + agency fees + creative production + sales/SDR cost + tools + onboarding cost attributable to acquisition) / Total new customers acquired

Choose the numerator based on what you want to manage: channel efficiency (media-only) vs business unit economics (fully-loaded).

Interpretation

• Lower CAC is generally better, but only relative to customer value and payback time.
• Payback mindset: a CAC of $200 may be great if you recover it in 2 months, and risky if you recover it in 18 months.
• Scope matters: CAC can be calculated per channel, per campaign, or blended across all acquisition.

Step-by-step: how to calculate CAC cleanly

1. Pick a time window (e.g., last month) and keep it consistent across spend and customers.
2. Define “new customer” as first-time purchasers (not leads, not sign-ups) for CAC.
3. Choose scope: media-only vs fully-loaded; channel-level vs blended.
4. Align attribution logic for counting customers (e.g., last-click, first-click, or modeled). Use the same logic every time you trend CAC.
5. Compute CAC and then compute payback (see below) to interpret it.

Numeric example

In April:

Continue in our app.

You can listen to the audiobook with the screen off, receive a free certificate for this course, and also have access to 5,000 other free online courses.

Or continue reading below...

Download the app

• Paid media spend: $30,000
• Agency + creative: $6,000
• Sales commissions attributable to new customers: $4,000
• New customers acquired: 200

CAC_media = 30,000 / 200 = $150 per customer

CAC_fully_loaded = (30,000 + 6,000 + 4,000) / 200 = 40,000 / 200 = $200 per customer

Payback check (simple): if average contribution margin per customer per month is $50, then:

Payback months ≈ CAC_fully_loaded / monthly contribution margin = 200 / 50 = 4 months

Common pitfalls

• Mixing scopes: comparing a media-only CAC this month to a fully-loaded CAC last month makes trends meaningless.
• Counting leads as customers: CAC becomes artificially low if the denominator is not “new customers.”
• Time-lag blindness: spend happens now, customers convert later. A short window can inflate CAC during ramp-up and deflate it during harvest.
• Ignoring fixed acquisition costs: creative, tools, and sales time can be material, especially in B2B or high-touch funnels.

Misread alerts

• CAC “improves” because you cut sales support: media CAC looks better, but close rates drop next month.
• CAC “worsens” during testing: early experiments often raise CAC temporarily; judge with learning goals and longer windows.
• Blended CAC falls due to brand demand spike: organic/direct customers rise, lowering blended CAC even if paid efficiency is unchanged.

2) Return on Ad Spend (ROAS)

What it measures

ROAS measures revenue generated per dollar of advertising spend. It is a revenue efficiency metric, not a profit metric—so it must be interpreted in the context of margins and costs beyond media.

Formula

ROAS = Attributed revenue / Ad spend

Two common variants:

• Gross ROAS: uses gross revenue in the numerator.
• Contribution ROAS (margin-aware): uses contribution margin dollars (or profit proxy) in the numerator.

ROAS_contribution = Attributed contribution margin / Ad spend

Interpretation

• ROAS = 4.0 means $4 of attributed revenue per $1 spent.
• ROAS is most actionable when you know your break-even ROAS based on margin and other variable costs.
• High ROAS can still be bad if it comes from heavy discounting or low-margin products.

Step-by-step: add margin context (break-even thinking)

1. Compute gross ROAS for the campaign/channel.
2. Estimate contribution margin rate (after COGS, shipping subsidies, payment fees, returns—whatever is relevant to your business).
3. Convert revenue to contribution margin dollars to see if ads generate enough margin to cover spend.
4. Compare to target (e.g., “Contribution ROAS must be > 1.0” or “Gross ROAS must be > 3.0”).

Numeric example

A campaign spends $10,000 and is attributed $50,000 in revenue.

ROAS = 50,000 / 10,000 = 5.0

If contribution margin rate is 30%, then contribution margin dollars are:

Attributed contribution margin = 50,000 * 0.30 = 15,000

ROAS_contribution = 15,000 / 10,000 = 1.5

This is much more informative: you generated $1.50 of contribution margin per $1 of ad spend.

Common pitfalls

• Short-term bias: ROAS often over-rewards campaigns that capture existing demand (e.g., branded search) and under-rewards campaigns that create future demand (e.g., prospecting).
• Attribution inflation: last-click attribution can over-credit bottom-of-funnel touchpoints, making ROAS look better than true incrementality.
• Revenue ≠ profit: ROAS can rise while profit falls if discounts increase or product mix shifts to low-margin items.
• Comparing ROAS across channels without context: different channels play different roles in the funnel and have different lag times.

Misread alerts

• ROAS up due to discounting: revenue may rise, but margin dollars can fall. Always check contribution ROAS or margin per order.
• ROAS up because you narrowed targeting: you may be harvesting high-intent users while starving top-of-funnel growth.
• ROAS down after creative refresh: short-term ROAS can dip while the system relearns; validate with longer windows and holdout tests when possible.

3) Lifetime Value (LTV)

What it measures

LTV estimates the total value a customer generates over time. The key is to calculate LTV in a way that matches your business model (subscription vs repeat purchase) and decision (budgeting, bidding, or forecasting).

Core approaches and formulas

Revenue LTV (simpler, but can mislead if margins vary):

LTV_revenue = Sum of expected customer revenue over a defined horizon

Margin (contribution) LTV (better for unit economics decisions):

LTV_margin = Sum of expected customer contribution margin over a defined horizon

Cohort-based LTV (recommended): track customers acquired in the same period and measure their cumulative value over time.

Cohort LTV at month t = (Cumulative revenue or margin from cohort through month t) / (Customers in cohort)

Subscription shortcut (when applicable):

LTV_margin ≈ ARPU * gross margin % / churn rate

Use shortcuts carefully; cohort measurement is usually more trustworthy.

Interpretation

• LTV is not a single “true number.” It depends on time horizon (e.g., 90-day, 12-month, 24-month) and whether you use revenue vs margin.
• Use LTV to decide how much CAC you can afford and how aggressively you can scale.
• When LTV is uncertain, prefer shorter-horizon LTV (e.g., 90-day margin LTV) for operational decisions.

Step-by-step: cohort LTV calculation

1. Create acquisition cohorts (e.g., customers whose first purchase occurred in January).
2. Track cumulative revenue (or margin) per cohort by month since acquisition.
3. Divide by cohort size to get per-customer cumulative LTV at each month.
4. Compare cohorts (by channel, campaign, product, or geography) to see which acquisition sources produce higher-quality customers.
5. Choose a horizon that matches decision speed (e.g., 6-month LTV for budgeting, 12-month for strategy).

Numeric example

You acquire 100 customers in January (one cohort). Their cumulative revenue is:

If contribution margin rate averages 40%, then 3-month margin LTV is:

LTV_margin_3mo = 150 * 0.40 = $60

When to use LTV:CAC

LTV:CAC is a ratio used to assess whether customer economics are healthy.

LTV:CAC = LTV (preferably margin-based) / CAC (preferably fully-loaded)

Use it when:

• You have a repeat purchase or subscription model where value accrues over time.
• You can estimate LTV with reasonable stability (cohorts have matured enough).
• You need a high-level guardrail for scaling acquisition.

Avoid over-relying on LTV:CAC when LTV is mostly projected rather than observed.

Common pitfalls

• Using revenue LTV for profit decisions: if margins vary by product or channel, revenue LTV can push you to overspend.
• Over-projecting early cohorts: extrapolating month-1 behavior into “lifetime” can dramatically overstate LTV.
• Ignoring retention drivers: LTV changes can come from retention, purchase frequency, AOV, returns, or product mix—don’t treat it as a black box.
• Mixing cohorts: averaging all customers hides whether new acquisition quality is improving or deteriorating.

Misread alerts

• LTV “increases” due to price hikes but churn rises: short-term LTV curves may look better before retention damage appears.
• LTV “improves” because only best customers are being acquired: targeting may be too narrow, limiting scale even though LTV looks great.
• LTV looks flat because horizon is too short: some channels have slower payback; compare cohorts at the same maturity (e.g., month-6 vs month-6).

4) Conversion Rate

What it measures

Conversion rate (CVR) is the percentage of users (or sessions, clicks, leads) that complete a defined action. The most important skill is choosing the right funnel stage and keeping numerator/denominator consistent.

Formula

Conversion rate = Conversions / Total opportunities

Examples of “opportunities” (denominator) depending on the stage:

• Click-to-purchase CVR: Purchases / Ad clicks
• Session-to-purchase CVR: Purchases / Website sessions
• Lead-to-customer CVR: New customers / Leads
• Checkout completion rate: Orders / Checkout starts

Interpretation

• CVR is a rate, so it can improve even if total conversions fall (if traffic drops faster than conversions).
• CVR is highly sensitive to traffic quality, device mix, geography, and intent.
• Use CVR to diagnose funnel friction, but always pair it with volume metrics (sessions, leads, spend).

Step-by-step: choose the right conversion rate

1. Pick the decision you’re making (landing page optimization, checkout fixes, lead qualification, ad targeting).
2. Select the funnel stage closest to that decision (e.g., checkout completion for payment issues; session-to-lead for landing page clarity).
3. Define numerator and denominator precisely (event names, deduping rules, time window).
4. Segment before you act (new vs returning, mobile vs desktop, channel, campaign, geography).
5. Trend and annotate changes (site changes, pricing changes, campaign launches) so you don’t attribute CVR shifts to the wrong cause.

Numeric example

In a week:

• Ad clicks: 20,000
• Website sessions: 15,000
• Purchases: 450

Click-to-purchase CVR = 450 / 20,000 = 2.25%

Session-to-purchase CVR = 450 / 15,000 = 3.0%

Both are “conversion rates,” but they answer different questions. If click-to-purchase drops while session-to-purchase stays flat, the issue may be click quality (or tracking differences between clicks and sessions), not onsite performance.

Common pitfalls

• Wrong funnel stage: using session-to-purchase CVR to judge ad creative can be misleading if the real issue is click quality.
• Inconsistent numerator/denominator: counting purchases by order ID but sessions by users, or mixing time zones/windows, can distort CVR.
• Not deduping conversions: double-counted purchases inflate CVR and hide problems.
• Ignoring mix shifts: more mobile traffic can lower CVR even if the site is unchanged.

Misread alerts

• Conversion rate up due to traffic quality changes: if you paused broad targeting, CVR may rise while total new customers fall.
• Conversion rate up because volume dropped: fewer sessions from low-intent users can raise CVR without improving revenue.
• Conversion rate down after expanding geos: CVR decline may be expected; evaluate by segment and adjust expectations.