Should companies optimize acquisition spending on CAC or on ROI (cohort quality)?

Martin Gontovnikas · Feb 14, 2024

Spending heavily on ads to drive sign-ups without checking whether those sign-ups activate is wasted money — he burned $300k producing a nice board number that generated no pipeline.

Almost none of the sign-ups activated, so they weren't retained and never converted; the metric was optically good but commercially empty.

Scope: happened early in the company's life

60:15 20Growth: How to Master Product-Led-Growth, The Biggest Mistakes Startups Make When Scaling into Enterprise, How to Assess "Bets" in Growth; Which to Take and Which to Not with Gonto, Interim CMO @ Vercel

Antoine Le Nel · Oct 18, 2024

Companies should optimize acquisition on ROI rather than CAC, because targeting CAC drives you to acquire the worst cohorts

Optimizing for a low cost per acquisition selects for the cheapest, lowest-quality users rather than the most valuable ones

Scope: describes how Revolut frames acquisition internally

0:00 20Growth: Revolut's Chief Growth Officer on The Growth Playbook Revolut Used to Scale to $2.2BN in Revenue | How Revolut Launch and Grow Products | Why the Best PMs Don't Need A/B Tests & Why CAC is a BS Metric with Antoine Le Nel

Antoine Le Nel · Oct 18, 2024

CAC is an irrelevant metric and teams should optimize ROI instead; he raised Revolut's CAC significantly while significantly improving ROI

If you set CAC targets you always acquire the worst, cheapest cohorts and go downhill — you'd only run cheap channels like TikTok even though the cohorts there are worse than from Instagram; you're not there to get the cheapest users but the best ones, and sometimes you have to pay more

Scope: requires granular channel-level cohort analysis

24:16 20Growth: Revolut's Chief Growth Officer on The Growth Playbook Revolut Used to Scale to $2.2BN in Revenue | How Revolut Launch and Grow Products | Why the Best PMs Don't Need A/B Tests & Why CAC is a BS Metric with Antoine Le Nel

Adam Grenier · Mar 15, 2023

CAC remains a useful operating input for media buyers as long as it is fed by a media mix model optimized for confidence rather than perfection

Telling a buyer 'we believe a Facebook customer is worth $20, go get as many as possible' lets them operate, rather than asking them to decipher the magic behind CAC-to-LTV themselves

Scope: requires pulling in more data, more models and partner information

44:35 20Growth: The Inside Story to Uber's Hypergrowth Scaling; What Worked, What Did Not? | Spending a $1BN Budget at Uber and Why China was the Wild West for Uber | Why You Do Not Need a Growth Team with Adam Grenier

Adam Grenier · Mar 15, 2023

Imperfect attribution doesn't remove the need for CAC as a buying tool, because buyers still have to decide where and how much to buy

A mobile gamer still needs to acquire as many $1 users as possible and must choose a channel and price to do it

46:04 20Growth: The Inside Story to Uber's Hypergrowth Scaling; What Worked, What Did Not? | Spending a $1BN Budget at Uber and Why China was the Wild West for Uber | Why You Do Not Need a Growth Team with Adam Grenier

Matteo Franceschetti · Nov 20, 2023

Founders should set a hard blended CAC cap and refuse to exceed it regardless of market conditions

The cap can be derived top-down from benchmarks (top companies' marketing spend as a share of revenue) and from the contribution margin you want — if you want 20% net margin from day zero, that dictates the CAC you can afford

12:33 Hard cac cap tied to contribution margin target should govern spend

Harry Stebbings · Mar 15, 2023

Media spend should be evaluated as LTV against audience size and probability of conversion, not via attributed CAC

If a three-month campaign costs $45k and a single $50k enterprise conversion out of millions of impressions is highly likely, the deal clears; if you'd need 50,000 conversions at a dollar each, it doesn't

45:26 Evaluate media spend via ltv against audience size and conversion probability not attributed cac

Nigel Morris · May 24, 2021

Unit economics are the absolute building block of an investment heuristic, because successive customer-level cohorts add up to the company's P&L

At Capital One the decision-making algorithm was built entirely on unit economics, joining customer-level cohorts to the company P&L; and each passing week yields more data and clarity

25:10 Unit economics are the fundamental building block of investment decisions since customer cohorts aggregate into the companys pl

Antoine Le Nel · Oct 18, 2024

Because spend is scaled gradually and good analytical models are built, you can know very quickly whether the cohorts you're acquiring are good ones

They scale spend gradually, and the models compound — he knows far more now than one or two years ago about where the good cohorts are and how much he's willing to pay, which is why growth is exponential

25:52 Gradual scaling plus analytical models reveals cohort quality quickly

Jason Lemkin · Jan 13, 2023

CAC is a deeply flawed metric because almost every individual marketing program looks like bad ROI on a per-deal basis.

With a $5k ACV you'll never spend $5k to acquire a $5k customer — some acquisition is free and viral, and paid programs cost more than first-year revenue; the math only works over multi-year retention, referrals, competitive wins and hitting plan.

15:45 Cac looks like bad roi per program but multiyear retention and referrals justify spend