School Me MarketingSMM
FoundationsIntermediate11 min deep read · 30-sec skim available

Marketing Metrics: CAC, Conversion & ROAS

Three ratios run modern marketing: what a customer costs, what share of lookers become buyers, what a rupee of ads returns. And each one lies fluently when read alone. Metric literacy isn't knowing the formulas; it's knowing what each number hides and which question it's allowed to answer. You can't bank a ratio.

⚡ Understand it in 30 seconds

  • The metric tree is one chain of arithmetic: impressions → clicks (CTR) → sessions → orders (conversion rate), with spend ÷ new customers = CAC and revenue ÷ spend = ROAS. Every dashboard number is a rung on this ladder, and every rung has a known failure mode.
  • CAC lies by blending: paid and organic mixed (blended CAC flatters), new and returning customers mixed (returning buyers make acquisition look cheap), and last quarter's brand effects paying for this quarter's 'performance'.
  • ROAS lies twice: it counts revenue (not margin), and it claims credit (attribution) rather than proving cause (incrementality). A 4x ROAS on 30% margins is barely break-even. A 10x ROAS on customers who'd have bought anyway is a subsidy with a dashboard.
  • Conversion rate lies by mix: improve traffic quality and it rises with no site change, or run a sale and it spikes while margin bleeds. Ratios are diagnostics, not goals: optimise a ratio in isolation and it will happily shrink the business to improve itself.

Go deeper

The core idea

Build the intuition on the chain before trusting any single link. A D2C brand spends ₹10 lakh on ads: 50 lakh impressions, 1% CTR → 50,000 sessions, 2% conversion → 1,000 orders, of which 700 are new customers. CAC = ₹10,00,000 ÷ 700 ≈ ₹1,429. Average order ₹2,000 → revenue ₹20 lakh → ROAS 2.0x. Looks alive, until margin enters: at 35% contribution, those orders earn ₹7 lakh against ₹10 lakh of spend. This '2x ROAS' business loses ₹3 on every ad rupee until repeat purchases arrive, which is why CAC's only meaningful reading is against contribution margin and repeat behaviour (payback period and CLV), never against revenue. The chain forces the discipline: every ratio is a numerator and denominator you must be able to name, and most metric arguments are secretly denominator arguments.

CAC's honest versions come from splitting. Paid CAC (paid spend ÷ customers attributed to paid) versus blended CAC (all spend ÷ all new customers): the gap between them is your organic engine, and watching paid CAC rise while blended holds is the classic sign you're buying customers the brand was already earning. New-versus-returning splits stop the flattering error of counting a loyalist's fifth order as an acquisition win. And the honest ceiling on all of it is incrementality: the customers who would have arrived anyway (they searched your name, and the discount intercepted them at checkout) make any attributed CAC look better than the causal one. The gold-standard check is the holdout: pause or geo-split the spend and watch what actually changes. Companies that run this test regularly are routinely humbled by how much 'performance' was reallocation of the inevitable.

ROAS needs three corrections before it can be trusted with a budget. Margin: convert to profit-on-ad-spend or set ROAS targets per category margin, since a fixed 4x target across a catalogue where margins run 15% to 60% systematically overfunds the wrong products. Attribution: platform-reported ROAS is a claim by an interested party. Click-based models over-credit the last touch, and every platform counts the same conversion. MER (total revenue ÷ total marketing spend) is the blunt but unbluffable supervisor: when platform ROAS rises while MER falls, the platforms are taking credit, not creating customers. Diminishing returns: ROAS is an average, but budgets move at the margin. The first lakh targets your eager buyers, the tenth chases strangers, so the question for more budget is always marginal ROAS, which no dashboard shows by default and every scaling decision depends on.

Ratios also need volume and guardrails to mean anything. A CAC of ₹300 at a thousand customers a month and the same CAC at fifty thousand are different businesses: efficiency metrics divorced from scale reward shrinking to excellence ('our best-ever ROAS' on a tenth the spend is usually a smaller company, not a better one). The working setup pairs every optimisation metric with a guardrail: conversion rate with average margin per order (kill the discount-spike illusion), CAC with new-customer volume and payback months, ROAS with MER and contribution margin. And a share of judgment stays reserved for what these metrics structurally can't see: the brand effects that make next year's clicks cheaper, which short-term dashboards record as nothing happening.

The business case

Why marketers care

Budget allocation runs on these numbers, so their failure modes become spending mistakes at company scale: blended CAC hides a dying paid channel, revenue ROAS overfunds low-margin SKUs, last-click attribution starves the top of the funnel that fills the bottom.

Metric literacy is the marketer's seat at the money table. The one who says 'ROAS is 4' reports. The one who says 'marginal ROAS at the next lakh is 1.8 against a 2.9 break-even, so the budget should move to retention' decides. Same data, different job grade.

And the metrics are how growth stories get audited, by boards, acquirers, and public markets. India's D2C wave learned this publicly: valuation debates increasingly run on CAC trends, repeat rates, and contribution margins, not topline growth. Reading your own numbers the way an auditor will is cheaper before the audit.

2.0×

ROAS on ₹10 lakh of ad spend: looks alive

The worked example from the core idea: illustrative numbers

35%

Contribution margin: the correction ROAS forgets to make

− ₹3 lakh

What those ad rupees actually earned, before repeat purchases arrive

See it

The visual model

The metric tree

One chain of arithmetic from impression to margin: every dashboard number is a rung, and each rung has a known way of lying.

Read this diagram as text

A funnel of six stages mapping metrics to the customer journey. Impressions: reach times frequency, where the spend enters. Clicks: click-through rate, the ad's promise measured. Sessions to carts: site conversion, polluted by traffic mix. Orders: where CAC is computed, spend divided by new customers, honest only when paid and organic, new and returning, are split. Revenue: where ROAS is computed, revenue divided by spend, a claim until margin and incrementality correct it. Margin and repeat: contribution margin, payback, and CLV, the bottom of the tree, where the truth lives.

The receipts

Where it comes from

The accountability tradition starts with Claude Hopkins (Scientific Advertising, 1923) and direct-response practice; digital analytics industrialised it; the corrective layer comes from incrementality research and Binet & Field's effectiveness work (The Long and the Short of It, 2013).

Hopkins' keyed coupons were CAC and conversion tracking with 1920s technology, the founding claim that advertising is a measurable investment, not an art subsidy. Digital didn't invent performance marketing. It dropped the cost of Hopkins' discipline to near zero and multiplied the places it could mislead.

The incrementality corrective formalised the difference between attribution (assigning credit among touchpoints) and causation (what changed because the ad ran). Holdouts, geo-experiments, and ghost ads exist because attributed conversions systematically include the inevitable, and the gap between attributed and incremental is largest exactly where ROAS looks best: branded search and retargeting.

Binet and Field supplied the time dimension: activation metrics (CAC, ROAS, conversion) capture short-term effects and structurally miss brand-building's compounding, which shows up later as cheaper acquisition and price tolerance. Their practical warning fits this page in one line: optimising wholly to efficiency ratios shifts spend toward what measures well this quarter, which is not the same as what grows the business.

Brands you know

Seen in India

Educational readings of familiar brands — how the concept helps you see what they do, not claims about their current campaigns.

Mamaearth

The D2C beauty house whose IPO put creator-and-ad-led growth under a public lens.

The listing-era debate around Honasa read as a national seminar on this page's metrics: commentary centred on advertising intensity, the economics of influencer-driven acquisition, and whether repeat behaviour and margins would let marketing spend taper, the CAC-versus-CLV question asked with a ticker attached. Whatever one's view of the stock, the episode showed which numbers the market actually audits: contribution margin, repeat rates, and the trajectory of spend-to-revenue, not follower counts.

What to steal: Build the dashboard you'd want read aloud at your IPO — growth bought above contribution margin is a story with a disclosure date.

MakeMyTrip

A survivor of Indian travel's great discount wars.

The OTA battles of the mid-2010s read as ROAS illusion at category scale: coupon-fuelled bookings produced spectacular topline and grim unit economics, as rivals effectively bought the same price-sensitive travellers from each other each quarter. The later era (consolidation, reduced discount intensity, brand strength carrying organic traffic) reads as the industry graduating from revenue metrics to margin-and-repeat metrics, the tree read from the bottom.

What to steal: When competitors subsidise the same switchers, everyone's acquisition metrics improve while everyone's business worsens — check whether your CAC is buying customers or renting deal-hunters.

Paytm

The fintech giant's public march from growth-at-any-CAC to contribution-margin discipline.

Its post-listing narrative arc (cashback intensity giving way to explicit contribution-profit targets and quarterly commentary on payback) reads as a company re-founding itself on the bottom of the metric tree. The cashback era optimised acquisition volume. The discipline era prices every incentive against the margin and repeat behaviour it purchases, in public, under analyst questioning.

What to steal: Incentives are CAC by another name: account for cashbacks, discounts, and freebies inside acquisition cost, or your CAC is a fiction with a marketing budget.

Beyond India

The global lens

Casper

The D2C mattress pioneer whose S-1 became a unit-economics cautionary tale.

The IPO filing revealed the arithmetic the brand's buzz had obscured: heavy per-customer acquisition and service costs against a purchase that recurs roughly once a decade, so CAC is repaid slowly, if ever, in a category with structurally low frequency. It's the canonical reminder that metric health is category-relative: numbers that thrive on repeat frequency cannot be willed onto an infrequent purchase.

What to steal: CAC's affordability is set by frequency and margin, not ambition — some categories can never out-market their own repeat rate.

Uber

The famous audit of its performance-ad spend.

The widely reported episode: investigating ad fraud, the company cut a nine-figure chunk of app-install spend, and installs barely moved. The attributed conversions had largely been organic installs wearing paid credit (and some outright fraud), making it the standing citation for incrementality testing: the metric said the spend worked, but the holdout said the spend was a donation.

What to steal: Attribution is testimony from interested witnesses. Periodically pause something big enough to matter and let reality testify instead.

From theory to Monday morning

How to use it

  1. Draw your own metric tree

    One page, your funnel's actual numbers at each rung, updated monthly. Most metric confusion dies when the chain is visible: everyone argues about ROAS until the tree shows the conversion rung (or the margin rung) is where the problem lives.

  2. Split CAC three ways before quoting it

    Paid vs blended; new vs returning; by channel with incentives (discounts, cashbacks) included in the numerator. Report the spread, not the average: the flattering blend is how dying channels hide inside healthy dashboards.

  3. Convert ROAS to margin terms

    Compute break-even ROAS per category (1 ÷ contribution margin) and judge every campaign against its own line: a 3x on 60% margin beats a 5x on 15%. Track MER beside platform ROAS, and when they diverge, believe MER and audit the attribution.

  4. Run one incrementality test per quarter

    Geo-holdout or audience holdout on your biggest 'best-performing' line item, with retargeting and branded search first, because they're where inevitability hides. Price the gap between attributed and incremental, and re-set channel budgets on the causal number.

  5. Pair every target with its guardrail

    Conversion rate with margin per order; CAC with new-customer volume and payback months; ROAS with MER. Put the pairs on the same chart: a ratio allowed to travel alone will improve itself at the business's expense.

Watch out

Common mistakes

✕ Reporting blended CAC while paid CAC deteriorates.

Fix: The blend borrows the brand's organic customers to flatter the ads. Split them. The paid line's trend, not the blend's level, is the early warning that scaling has hit the ceiling.

✕ Treating platform ROAS as ground truth.

Fix: Platforms grade their own homework, and several claim the same conversion. Reconcile against MER and holdout tests. Budget on incremental returns, and expect the correction to be largest where reported ROAS is proudest.

✕ Optimising ratios without volume floors.

Fix: Cutting to the best-performing sliver produces beautiful ratios on a shrinking base. Set the volume floor first (customers, margin rupees), then optimise efficiency within it. The goal is the most profit, not the prettiest division.

✕ Judging brand spend on activation dashboards.

Fix: Brand effects arrive as next year's cheaper CAC and stronger conversion, invisible to this quarter's ROAS. Evaluate brand work on its own instruments (trackers, branded search trend, base sales) or the dashboard will quietly defund your future demand.

Don't just read it

Practice task — 10 minutes

Take any campaign you can see numbers for: yours, a client's, even a public case. Rebuild its metric tree: spend to impressions to clicks to orders to margin. Compute paid CAC (incentives included), break-even ROAS at its real margin, and mark which single number in the chain you'd holdout-test first. Then write the sentence the dashboard was hiding.

If you remember five things

  • The metric tree is one chain, impressions to margin, and every ratio is a rung with a named numerator, denominator, and failure mode.
  • CAC honest = split (paid/blended, new/returning, incentives in) and read against contribution margin and payback, never revenue.
  • ROAS needs three corrections: margin (break-even = 1 ÷ CM), attribution vs incrementality (holdouts decide), and marginal-not-average returns for scaling.
  • Ratios travel with guardrails and volume floors: efficiency alone rewards shrinking the business to look brilliant.
  • Short-term dashboards structurally miss brand compounding. Keep separate instruments for the spend that makes next year cheaper.