Lachlan Todd
The founders we talk to are sophisticated, but they still get this wrong.
CPM, cost per mille, is the cost to get 1,000 views on a piece of content. A video that costs $10 and gets 1,000 views has a $10 CPM.
It sounds like a useful number, but here is why chasing a low CPM is one of the most expensive mistakes you can make.
1. Views are not installs
Clipping agencies remix existing content, get paid by views, and typically charge $2-3 CPM. Sounds like a deal.
The problem: a huge percentage of those views come from users who will never buy your product. Low-income markets, wrong demographics, people who liked a trending audio clip. None of that becomes a customer.
Some UGC agencies do the same thing with original content. They optimize for view count using formats that go viral but never demo the product. Text-wall videos, reaction content, POV humor. Viral, yes. Zero conversion.
If you spend $20,000 and get 10 million views at a $2 CPM, but those views produce 200 installs, your real cost per install is $100. A campaign with a $15 CPM that produces 5,000 installs costs $4 per install. Anyone should prefer the second outcome.
CPM tells you nothing about which one you got.
2. CPM ignores the cost of finding winning creatives
Say your best ad runs at a $10 CPM. Looks clean.
But now count the full cost. Each ad costs $50 to produce. Meta's own published figure is roughly 1 in 25 ads becomes a winner. That means you spent $1,250 in production to find that one winner. Add agency or in-house overhead and you might spend $3,750 a month to find it.
At $10K/month in ad spend, you are really spending $15K per month total. Your true CPM is $15, not $10.
At $50K/month, it moves less, but the absolute spend to find winners is higher. The point is the same: the cost of the creative you threw away to find the winner is invisible in your CPM number. CAC captures it. CPM does not.
3. A low CPM usually means you are reaching the wrong people
If you run completely untargeted content, your CPM will be very low. Broad reach, low cost per thousand, great-looking number.
If you want to reach 28-38 year old women with anxiety who are already searching for mental health apps, that is a harder audience to find. You will pay more per thousand impressions to reach them.
Founders chasing the lowest possible CPM often end up unconsciously optimizing to reach everyone, which in practice means reaching no one who will buy. The cost per view goes down. The cost per install quietly goes up.
CAC captures the full picture. What did you spend to get a paying user? Everything else, CPM included, is just one variable inside that answer.
What to optimize for instead
The only number that tells you whether your marketing is working is Customer Acquisition Cost relative to what a customer is worth.
There are 3 components that we focus on for this:
Organic: conversion-focused content that converts around ~0.3% (depends on exact project). At our starting point of 2M views monthly, that's ~6,000 users minimum.
Paid Ads: we highly recommend running paid ads as well. Organic delivers 500+ creatives minimum, which often has ~5-20 winning ad creatives and drops CAC over 10%. At $100k/mo in ad spend, that's $10-20k in additional revenue.
Funnel Optimization: an additional 10% in traffic means nothing if your backend is a leaky bucket. Optimizing every step along your user's onboarding journey, including your email sequences for those that didn't purchase, can easily outweigh this. A 5% fix to 5 different drop-off points compounds to a 27.6% boost.
This is why we measure our results in CAC reduction, not CPM. Our goal is to flood your ad account with enough high-quality, conversion-engineered creatives that your blended CAC comes down. Across our clients, that is typically 20% or more, often within the first 90 days.
