CJ Gustafson published a good piece with AJ Ljubich from Datadog's FP&A team arguing that CAC payback is more useful than LTV:CAC. The operational argument is right. LTV:CAC moves too much on assumptions to run anything with, and payback is a number you can act on this quarter.
I want to point at the formula rather than the argument.
What's in the denominator
They define it as sales and marketing costs divided by revenue additions, times gross margin. And revenue additions is the year-on-year change in the most recent quarter, annualized.
That's net new revenue. All of it. Which means expansion is sitting in that denominator with no acquisition cost attached to it.
Run the same $10M of S&M at an 80% margin three ways.
$8M of additions, all new logos. 12 month payback.
$4M new logos and $4M expansion. 12 month payback.
$2M new logos and $6M expansion. 12 month payback.
Three different businesses. One number.
Why that matters more than it looks
A new-logo dollar costs somewhere between 35 and 50 cents to acquire. An expansion dollar runs in the low teens. Renewals are single digits.
Blend them and you get a metric that reads as acquisition efficiency while being substantially a report on how well your existing customers are growing.
Two things follow and the second one costs money.
A company with a working expansion motion looks like it has efficient acquisition. It doesn't. It has efficient expansion, filed under the wrong heading, and the acquisition machine underneath is doing worse than the number says.
A company with weak expansion looks like it has an acquisition problem. So it spends more on sales and marketing, which is the one move that can't fix what's actually wrong, and the number gets worse, and the response to that is usually more of the same.
The part I'd have led with
The piece is framed as payback instead of LTV:CAC. I don't think they're alternatives, because the fastest way to improve payback is a lifetime-value activity.
$12,000 to acquire a customer. They pay $1,000 a month at an 80% margin. Payback lands at month 15.
Get that customer to expand to $1,500 at month three and payback lands at month 11. Four months earlier, 27% faster, on identical acquisition spend. Do it at month six instead and you still take three months off.
That expansion didn't only raise lifetime value. It moved the payback date.
Which is the thing that gets lost when people call LTV slow. It is slow. If you're doing this well you won't know what a customer is really worth for years, because they keep staying and keep growing. That's an argument for not waiting on the number. It isn't an argument for ignoring what produces it.
What to do with it
Split the denominator. Payback on new-logo revenue tells you what acquisition bought. Payback including expansion tells you what the account actually did.
Run both. The distance between them is the value of everything that happens after the sale, and right now it's invisible in either number on its own.
There's a longer version of that argument, including why excluding expansion is defensible and what it costs you, which is worth its own piece.
Lincoln Murphy formally named and popularized Customer Success starting in 2010 and has spent 15 years connecting it to expansion revenue and commercial outcomes. Read The Premise.