Review
Growth Units does one thing and does it well: it teaches you to actually calculate Customer Acquisition Cost and Lifetime Value, not just recognize the acronyms. It reads closer to a tight explainer than a full book, but that's the point. If you've ever nodded along in a meeting when CAC or LTV came up without being able to do the maths yourself, start here.
Key Takeaways
The parts worth keeping:
Customer Acquisition Cost
- Total marketing spend divided by new customers in the period feels like CAC, but it's not useful: it ignores people who convert later, it doesn't tell you which activities actually worked, and it gives you nothing to act on.
- The useful version: CAC = cost to get a potential customer in the door, divided by the conversion rate to becoming a customer.
- Break CAC out by channel wherever you can. A blended number hides which channels are actually working and which need to be tweaked or dropped.
- Watch for saturation: push too hard on one channel and only the harder-to-convert prospects are left, which quietly drives up your CAC.
A conversion funnel makes the drop-off visible at each stage, from someone landing on the site through to becoming a paying, referring customer:
| Stage | What it means | Typical loss |
|---|---|---|
| Acquisition | Getting someone onto the site | 0% |
| Activation | Sharing an email, signing up | 30% |
| Retention | Staying engaged for some time | 21% |
| Referral | Telling others | 64% |
| Revenue | Becoming a paying customer | 75% |
Lifetime Value
- Price multiplied by number of purchases looks like LTV, but skips the cost side entirely. The real formula is (price minus cost to produce) multiplied by number of purchases, the contribution margin times repeat purchases. Don't collapse it into one number, the components are worth seeing separately.
- A single static LTV figure is always going to mislead. Chart it as a flow over time instead, it shows you when you break even and what would actually move the number.
- Track LTV by cohort (customers grouped by the month they joined) rather than in aggregate. Lining cohorts up by their relative month, not calendar month, makes retention curves and revenue predictability far easier to read.
- Retention isn't automatically good news to lose. Negative churn happens when the customers who remain after a period spend enough more, through upgrades or repeat purchases, that revenue per customer goes up even though some customers left. It's also known as positive net revenue retention, and it's measured on revenue, not headcount.
Putting CAC and LTV together
- Acquisition Cost Ratio = LTV : CAC. Rule of thumb target is 3:1 or 4:1, a buffer that covers the rest of your costs, gives room for the payback period, and absorbs some risk to LTV itself. If the ratio drops too low, slow down growth spend rather than push through it.
- Steve Blank's ideal customer profile (from 4 Steps to the Epiphany) is a useful filter before you spend anything acquiring someone: they have the problem, they know they have it, they're actively looking for a solution, they've already hacked something together to solve it, and they can get budget to pay for a real one.
- Most of humanity doesn't have the problem your specific business solves. Spend accordingly.