Retention by Acquisition Channel Benchmarks
Breaking retention out by the channel that acquired the cohort exposes the paid source that looks profitable on CAC but loses money on the second order. Here's what good and bad look like by channel.
Retention by Acquisition Channel
Repeat-purchase and revenue retention measured separately for each acquisition channel, so you can compare customer quality, not just acquisition cost.
Retention by acquisition channel splits your repeat-purchase curves by the source that brought the customer in — paid social (Meta, TikTok), paid search (Google), organic search, direct, email/referral. Instead of one blended retention number, you get one curve per channel.
The point is diagnostic. Channels with low CAC often look like the winners in a blended P&L, but two customers acquired at the same cost are not worth the same money. A Meta prospecting customer who never buys again is not the same asset as an organic customer who comes back three times. This view is where that gap becomes visible.
Most stores discover they have a channel quality problem the same way: a paid channel hits its CAC target for six months, then blended LTV starts slipping and nobody can explain why. The blend is hiding the fact that one source is dragging the average down.
Splitting retention by channel turns that mystery into a measurable gap. You stop arguing about whether Meta prospecting is 'working' in the abstract and start comparing 90-day repeat rates against Google brand or organic — channels you already know convert high-intent buyers.
Repeat-purchase rate by acquisition channel — typical DTC apparel and beauty ranges (months after first order)
| Acquisition channel | M1 repeat % | M3 repeat % | M6 repeat % | M12 repeat % | LTV index (blended = 100) |
|---|---|---|---|---|---|
| Organic search | 12-16% | 26-32% | 38-46% | 52-60% | 135-150 |
| Direct / brand | 14-18% | 28-34% | 40-48% | 55-62% | 140-155 |
| Email & referral | 18-24% | 34-42% | 48-56% | 62-70% | 160-180 |
| Google paid (brand) | 11-15% | 24-30% | 36-44% | 50-58% | 125-140 |
| Google paid (non-brand) | 8-11% | 18-24% | 28-34% | 40-46% | 95-105 |
| Meta prospecting | 6-9% | 14-20% | 22-28% | 32-40% | 70-85 |
| TikTok prospecting | 5-8% | 12-17% | 19-25% | 28-35% | 60-75 |
| Influencer / affiliate | 7-10% | 16-22% | 25-31% | 35-43% | 80-95 |
The shape is more important than the exact numbers. Owned channels (organic, direct, email) anchor the top. Brand-intent paid sits just below. Cold prospecting on Meta and TikTok lands 30-45% below the blended average — which is why a channel can hit CAC targets and still lose money on a contribution-margin basis.
12-month repeat-purchase curve by acquisition channel
Email & referral
Organic search
Google non-brand
Meta prospecting
TikTok prospecting
How to read these numbers for your store
Start by pulling cohorts of at least 500 first-time orders per channel — anything smaller and channel-level retention is mostly noise. If a paid channel doesn't have that volume in a quarter, roll up Meta + TikTok into 'paid social prospecting' for the read.
Then pair the curves with CAC by Channel. A 35% M6 repeat rate on Meta prospecting at a €28 CAC is a different conversation than the same retention at €55 CAC. The diagnosis is always the ratio: contribution margin per acquired customer over 6-12 months, not retention in isolation.
The Meta-vs-organic gap is the one to watch
If your Meta prospecting M6 repeat rate is more than 20 percentage points below organic, you're acquiring a fundamentally different customer — discount-led, low intent, single-purchase. Adding more budget will scale the leak, not the business. Fix creative and offer first, or shift the spend toward retargeting and brand search where intent is already higher.
What to do once you have the diagnosis
Cap or reallocate spend on the worst-retaining paid source first. The fastest win is usually to move 20-30% of cold prospecting budget into retargeting and branded search, where customer quality is closer to organic. You'll often see blended LTV recover within two cohort months.
Then attack the offer. A first-order discount that's too aggressive on Meta is the single biggest retention killer — customers anchored at -25% rarely come back at full price. Test a free-shipping threshold or a bundle instead, and track whether the M3 repeat curve closes the gap to Google non-brand.
Frequently asked questions
LTV by channel rolls everything — repeat rate, AOV, margin — into one currency number per cohort. Retention by channel is the underlying behaviour: how often customers come back. You need both. Retention tells you what's wrong; LTV tells you how much it costs you.
For DTC apparel and beauty, a healthy Meta prospecting cohort lands at 14-20% M3 and 22-28% M6. Below that range you're likely paying for discount-hunters. Above it usually means strong creative-product fit or a non-discount offer.
Aim for 500+ first-time orders per channel per cohort. Below 200 the curves swing wildly month to month. If a small paid channel can't hit that, group it with similar channels (e.g. all paid social prospecting) for the analysis.
First-touch is the right default for retention analysis — you're asking which source actually introduced this customer. Last-touch will over-credit email and retargeting, both of which are converting customers acquired elsewhere.
Meta prospecting is interruption marketing — you're showing ads to people who weren't shopping. Google non-brand catches active intent ('best running shoes'), and Google brand catches existing demand. The lower the intent at acquisition, the lower the repeat rate.
Partially. You'll have enough volume to see organic vs paid as a whole, but splitting paid into Meta vs TikTok vs Google non-brand often won't reach statistical reliability. Stick to two or three rolled-up groups until cohorts hit 500 orders.
Retention curves vary significantly by region — a US Meta cohort often retains differently than an EU one because of shipping cost, payment friction, and competitive intensity. If you're on Shopify Markets, split the analysis by region first, then by channel within each.
They're the two halves of channel-level unit economics. CAC by Channel tells you what you paid; retention by acquisition channel tells you what you got. A channel only makes sense to scale when the ratio of LTV-from-retention to CAC stays above your contribution-margin target — usually 3:1 or better.
Monthly at minimum, with rolling 90-day cohorts. Creative cycles, offers, and iOS attribution shifts all change channel quality faster than most teams expect. A quarterly review will miss the window to reallocate budget before a bad quarter compounds.
Yes — once you have the channel-level retention data, the pattern recognition is mostly mechanical. Tools that import historical GA4 plus order data can flag which cohort dropped, when, and which on-site behaviours correlate with the drop, so you go straight to testable hypotheses instead of guessing.
Track CAC, channels, and funnel conversion in one place
Metricuno connects ad spend, funnel events, and revenue so you can see CAC by channel, cohort, and campaign — without stitching together five tools.