When does each channel break even and how does it compound?
№ 162 · ROAS curves by channel
Definition
MetricCumulative return on ad spend by channel, plotted against days since spend
Unitrevenue per unit of ad spend (x), or cumulative % of spend recovered for app cohorts
Multiple lines (one per channel) showing cumulative revenue divided by cumulative spend over days. Different channels mature differently (search ramps slow, social ramps fast then plateaus). Build per-channel cohort by tracking attributed revenue vs spend. Example: Search 1.0x at D30; Display 1.0x at D90.
- ROAS
- Return on ad spend. Revenue attributed to a campaign divided by the spend on that campaign.
- D30
- Day 30 retention. Share of a signup cohort returning on the thirtieth day after install or signup.
Benchmarks
| Bottom 30% | Median | Top 30% |
|---|---|---|
| Blended ROAS below 2.0x, which is roughly half the ecommerce population; curve still under 1.0x at day 30 | Blended ecommerce ROAS 2.87x, Meta 1.86x to 1.93x, Google 3.52x to 3.68x, TikTok 1.41x; app D7 ROAS 4% to 8% depending on monetization model | Blended ecommerce ROAS at or above 3.5x, with the curve crossing 1.0x inside the first purchase cycle; app D30 ROAS above 45% on iOS |
DTC and ecommerce brands on Shopify-class stacks, full-year 2025, platform-reported ROAS at the channel level (revenue attributed by the ad platform divided by that platform's spend), plus mobile app cohorts measured as cumulative revenue over cohort spend at D7 and D30 · Two different definitions sit inside this card. Ecommerce channel ROAS is a multiple (2.87x) and is platform-reported, so it double counts across channels and overstates the sum. App ROAS is a cumulative percentage of spend recovered by day N and will read under 100% for months even in healthy cohorts. Do not compare the two on one axis. Triple Whale's MER of 41% is spend over revenue, the inverse convention, and inverts to a blended 2.4x. Liftoff figures come via Segwise 2025 and are single-source per genre, so treat the app rows as directional.
| Category | Bottom 30% | Median | Top 30% |
|---|---|---|---|
| Meta (paid social, DTC) | below 1.3x | 1.86x to 1.93x | 2.5x and above |
| Google Ads (search and shopping, DTC) | below 2.2x | 3.52x to 3.68x | 5.0x and above |
| TikTok (paid social, DTC) | below 1.0x | 1.41x | 2.0x and above |
| Amazon Ads (retail media) | below 2.0x | approx. 3.0x, with Electronics at 3.98x | 4.0x and above |
| Mobile app, casual games, D30 | below 20% iOS | 47% iOS, 15% Android | 55%+ of spend recovered on iOS |
When it looks bad
Every channel curve flattens below the 1.0x breakeven line and the gaps between channels stay constant, meaning no channel is compounding and spend is simply buying a fixed, immediate return.
Meta plateaus at 1.4x by day 14 and does not move by day 90, Google sits at 2.1x flat, and blended MER holds at 48% while contribution margin is 42%, so the business loses money on every incremental dollar.
What to do about it
- Set the breakeven line on the chart at 1 divided by contribution margin rather than at a fixed 3.0x or 4.0x target; at 40% contribution margin the real breakeven is 2.5x, and at 60% it is 1.67x, which reclassifies channels that look weak on a generic benchmark.
- Split each channel curve into prospecting and retargeting before reading it; Meta retargeting in beauty runs near 3.50x against a prospecting-weighted vertical median of 1.57x, so a blended Meta line hides which half is actually failing (rule1, drawing on Triple Whale).
- Move the primary chart from platform-reported ROAS to MER against blended spend; Triple Whale brands saw MER improve 9.86% in 2025 on Google while platform ROAS fell 10.03%, which is the signature of attribution drift rather than real performance loss.
- For app cohorts, chart D7 against the D30 to D7 multiple rather than D7 alone; casual titles that recover 47% by D30 on iOS versus 15% on Android need separate platform bid targets, not one blended ROAS goal.
Sources
- Triple Whale Median AOV across paid channels $74.12, up 2.68% YoY; top blended-ROAS verticals Sports 2.85, Travel 2.81, Home & Garden 2.65, Automotive 2.08 triplewhale.com ↗
- Triple Whale Meta median ROAS 1.93, median CVR 1.57%, median CTR 2.19%; MER by vertical from Media & Publishing 0.72 down to Automotive 0.34 triplewhale.com ↗
- Triple Whale Google took 23.03% of Triple Whale brand ad spend in 2025 versus Meta 68.31%; ROAS fell 10.03% and CVR fell 9.28% while MER improved 9.86% triplewhale.com ↗
- Triple Whale Electronics leads Amazon ROAS at 3.98 but sits bottom on MER at 0.23; Health & Wellness leads MER at 0.40 triplewhale.com ↗
- Triple Whale Median MER for Triple Whale customers in 2025 was 41%, expressed as ad spend over total revenue triplewhale.com ↗
- rule1 Median ecommerce ROAS 2.87x, down 4% YoY; by platform Google Ads 3.52x, Meta 1.86x, TikTok 1.41x; half of ecommerce businesses run below 2.0x rule1.ai ↗
- Eightx Meta median ROAS 1.93x with vertical spread from Automotive 2.54x to Media & Publishing 1.17x; Google beats Meta in 8 of 9 paired verticals at a median 50% premium eightx.co ↗
- Liftoff Hyper-casual D7 ROAS approx. 7.6% iOS and 7.8% Android, mid-core 4.3% iOS and 6.1% Android; D30 ROAS for casual games averages 47% iOS and 15% Android liftoff.ai ↗
- AppsFlyer Ranks media sources on retention, conversion and ROAS; Google Ads and Apple Ads lead, with AppLovin, Mintegral, Meta and TikTok closing the gap in gaming and non-gaming appsflyer.com ↗
- AppsFlyer Live benchmark tool for cost, retention and revenue metrics segmented by app vertical, platform and geography appsflyer.com ↗