ROAS meaning vs payback period: which one actually tells you if your ads worked
If you sell something cheap, fast, and once — a t-shirt, a sticker, a one-off gadget — ROAS is a fine way to read your ads. Spend $1, get $3 back the same day, repeat.
If you sell anything with a cycle longer than the checkout — a subscription, a service, a product people buy again — ROAS will lie to you, confidently, for months. Payback period won't. Here's the difference and when each matters.
ROAS meaning
ROAS meaning, in one line: attributed revenue divided by ad spend. Return on ad spend: attributed revenue divided by ad spend, usually per campaign or per channel. A 3x ROAS means $3 of attributed revenue for every $1 spent.
ROAS is fast to compute, lives in every ad platform natively, and is a perfectly good steering signal for delivery. The problem is what it leaves out:
It ignores COGS. $3 of revenue on a product with 70% COGS is $0.90 of gross profit on $1 of spend. That 3x ROAS is a loss.
It ignores time. Revenue attributed to today's spend might arrive over the next year. ROAS reports it as if it's instant.
It depends on the platform's attribution. "Attributed" is doing a lot of work — see did your Meta ads actually pay back for why that number is softer than it looks.
What payback period is
The number of months it takes for the gross profit from a cohort of customers acquired by a channel to equal the cost of acquiring them. At 5 months payback, the fifth month's gross profit from that cohort pays off the spend that acquired it; everything after is profit.
Payback period answers the question ROAS fakes: "when does this spend become free?"
How to compute payback period
You need three inputs per channel:
CAC — total spend on the channel divided by new customers acquired through it.
Gross margin — revenue minus COGS, as a fraction of revenue.
ARPA — average revenue per account per month, for the customers the channel brings in.
Then:
payback (months) = CAC / (ARPA x gross margin)
A channel with $200 CAC, $60 monthly ARPA, and 80% gross margin pays back in $200 / ($60 x 0.8) = 4.2 months.
That's the honest unit of "did the ads work." If payback is under your churn horizon, the channel prints money. If it's longer than customers stick around, the channel burns it.
When to use which
Use ROAS when | Use payback period when |
|---|---|
Single-purchase products | Subscriptions and recurring revenue |
Short, observable cycles | Long cycles where revenue accrues over time |
Comparing campaigns within a channel | Comparing channels against each other |
You need a number the platform gives you for free | You need a number finance will believe |
Most teams need both. ROAS to steer the campaign day-to-day; payback to decide whether the channel earns its budget next quarter.
Why most teams don't compute payback
Not because it's hard — the formula is one line. Because the inputs are split across three systems: spend in the ad platform, revenue in the database, and "which channel acquired this customer" requires joining identity from first touch through to the order. That join is the actual work.
FunnelKeeper does the join for you — Meta and Google spend, GA4 funnel events, and conversions and revenue from your database, all keyed by identity — so payback comes out per channel and per cohort without you stitching it. See how it works.
FAQ
What's a good payback period? For SaaS, under 12 months is healthy, under 6 is strong. For ecommerce it's usually measured in weeks. The real benchmark is your churn horizon: payback should be shorter than how long the customer sticks around.
Can I use ROAS for subscription businesses at all? As a delivery signal, yes. As a decision number, no — it'll push you toward channels that look efficient in week one and bleed in month six.
What about LTV:CAC? It's the same idea stretched further out: total lifetime value vs. acquisition cost. Payback period is just the time-to-breakeven slice of it, and it's the slice you can actually measure early instead of guessing LTV.