Is AppLovin's ROAS Really Higher? What Triple Whale's 755-Brand Study Found

On September 3, 2026, Triple Whale published one of the largest independent looks at AppLovin performance so far: Does AppLovin Actually Drive Revenue? Here's What the Data Says, written by co-founder Maxx Blank. Triple Whale is an attribution and analytics platform used by Shopify brands, and it has offered an AppLovin integration since November 2025, so it sees AppLovin spend and orders next to every other channel for a wide set of stores.

The headline number spread quickly: brands running AppLovin saw a 2.90 ROAS on it, compared with 2.08 on the other platforms in the study. That is a big gap. But the more useful part of the report is how Triple Whale tried to check that number three different ways, and where it is careful about what the data can and cannot say. Here is a plain-English breakdown, plus what it means if you are deciding whether to test AppLovin.

The short version

• Attribution: across 755 ecommerce shops over 12 months, AppLovin returned 2.90 ROAS vs 2.08 for the other ad platforms. 61% of shops saw a higher ROAS from AppLovin than from their other channels.
• Media mix model: a model trained on about 6,300 shops found 66% of AppLovin shops got a higher incremental return per dollar from AppLovin than from other platforms.
• Geo holdouts: 5 of 7 real-world holdout tests showed a statistically significant revenue lift, from +3.5% to +13.5%. The other 2 were not significant.
• The catch: the edge showed up most for brands that gave AppLovin real budget and time. It is not a reason to cut channels that already work.

What Triple Whale studied

The report looked at 12 months of data from 755 ecommerce shops that spent on AppLovin. Across that group, AppLovin was still a small piece of the budget: just 7.7% of combined ad spend. That matters, because it means most of these brands were still spending the bulk of their money on the channels they already knew, and AppLovin was being judged as an add-on.

Instead of relying on one number, Triple Whale used three measurement methods, each with different blind spots:

1. Multi-touch attribution: which channel gets credit for each order.
2. A media mix model (MMM): a statistical model that estimates how much revenue each channel actually adds, not just which one was clicked last.
3. Geo holdout tests: turning spend down in some regions and comparing revenue against regions where it kept running.

Each method answers a slightly different question. Attribution tells you who got the credit. The MMM estimates what would have happened without the channel. Geo tests measure it directly, but only for one brand at a time. If you want the background on why these methods can disagree, we cover it in AppLovin attribution and incrementality.

Finding 1: 2.90 vs 2.08 ROAS on attribution

Using Triple Whale's attribution data, AppLovin's ROAS came in at 2.90, compared with 2.08 for the other platforms the same shops were running. That is roughly 39% more revenue credited per dollar of spend.

It was not just a few big winners pulling up the average. 61% of qualifying shops saw a higher ROAS from AppLovin than from their other ad channels. The flip side is that about 4 in 10 did not, which is worth remembering before treating the average as a forecast.

Triple Whale is upfront about the weakness here: attribution decides credit by clicks, and click-based models can hand a channel credit for a sale that another channel (often social) helped create. That is exactly why the report did not stop at attribution.

What a 0.82 ROAS gap means in margin terms

A quick way to read any ROAS number is against your breakeven ROAS, which is 1 divided by your gross margin after product cost, shipping and fees. A brand with a 60% margin breaks even at about 1.67 ROAS. A brand with a 40% margin needs 2.50. At 40% margin, the difference between 2.08 and 2.90 is the difference between losing money on first orders and making a profit on them. The lower your margin, the more a gap of this size matters. Just keep in mind these are attributed figures; the next two findings test how much of that revenue is truly incremental.

Finding 2: the media mix model agrees

To get past the click-credit problem, Triple Whale ran its Foundation MMM, trained on roughly 6,300 shops: the 755 AppLovin spenders plus a large contrast group of shops that did not use AppLovin. The model was checked against three months of data it had not seen, where it explained 84% of the variation in revenue (an R² of 84%).

The result: 66% of AppLovin shops got a higher incremental return per dollar from AppLovin than from other platforms. In plain terms, the model estimates that for two out of three brands, an extra dollar on AppLovin added more revenue that would not have happened anyway than an extra dollar elsewhere.

An MMM is still a model, built on assumptions, so it is best read as a second opinion rather than proof. The value here is that a method with very different blind spots from attribution pointed the same way.

Finding 3: seven geo holdout tests

The closest thing to a real experiment is a geo holdout: pull spend back in some regions, keep it running in others, and compare actual revenue. Triple Whale reported seven of these tests, run between July 2025 and June 2026, with spend held back per test ranging from about $16.7K to $152.1K.

Five of seven tests showed a significant lift, averaging 8.3% according to Triple Whale. Two did not. That is an honest result: AppLovin added measurable incremental revenue for most of the brands that tested it, but not automatically for every brand at every budget.

Does this mean AppLovin beats Meta?

This is how the study has been shared on social media, so it is worth being precise. The report compares AppLovin with the other platforms in the study as a group. It does not break the 2.08 benchmark out by channel, so it is not a clean AppLovin vs Meta, or AppLovin vs Google, number.

What it does show is that, for the brands in the study, AppLovin returned more per dollar than the mix of channels they were already running, while most of their budget still sat outside AppLovin. For most DTC brands that existing mix is built around Meta, which is why the comparison matters to them.

Triple Whale's own explanation for the gap is useful: less competition on a channel usually means stronger marginal returns, which it compares to what early TikTok advertisers saw before that channel got crowded. That is an argument for testing now, not a promise that the gap lasts. For a side-by-side of how the two platforms work, see AppLovin vs Meta ads.

What the study does not prove

Good data deserves a careful read. A few limits to keep in mind:

• Budget and patience matter. The report says results were strongest for brands that gave AppLovin real budget and runway, and warns against treating it as a 60 to 90 day toe-dip.
• Early-channel effect. Part of the advantage likely comes from AppLovin still being a small share of spend. Returns can compress as more advertisers arrive.
• Not every test was a win. Two of seven geo tests were not significant, a reminder that lift depends on the brand, the spend level and the test design.
• Selection. The 755 shops chose to spend on AppLovin. Brands that expected it to work, or that already had strong video creative, may be overrepresented.
• It is an average. A 2.90 average ROAS does not tell you what your brand will get. Your product, price point, creative and offer decide that.

Triple Whale's overall read is measured: AppLovin is promising and worth a real test, but not a reason to abandon what is already working. We agree with that framing.

How to test AppLovin on your own spend

Getting started is easier than it was a year ago. AppLovin opened its self-serve platform, now called AppLovin Ads, to all advertisers on June 22, 2026, after an eight-month referral-only phase, so you no longer need a referral code or a sales contact. The report's practical advice matches how we approach launches:

1. Fund it like a real channel. Commit enough budget and enough weeks for the algorithm to learn. Our AppLovin ads cost guide covers realistic starting budgets.
2. Avoid launching or measuring during BFCM week. Peak-season noise makes clean reads hard. If you want AppLovin live for Q4, launch before the rush; our BFCM playbook has the timeline.
3. Run your own geo test. Score the channel on incremental new-customer revenue against your social-channel baseline, not on platform ROAS alone.
4. Re-measure as you scale. Watch marginal returns (what the next dollar brings in), not just the average.
5. Feed it creative. AppLovin ads are full-screen video placements, so performance depends heavily on a steady supply of new video creative.

If you would rather have a team that does this every day handle the setup, testing and scaling, Apt Digital is an official AppLovin Partner Agency for ecommerce brands.

FAQ

What did the Triple Whale AppLovin study find?

Across 755 ecommerce shops over 12 months, AppLovin returned 2.90 ROAS compared with 2.08 on the other ad platforms those shops ran. 61% of shops saw a higher ROAS from AppLovin, a media mix model found 66% got a higher incremental return per dollar, and 5 of 7 geo holdout tests showed a significant revenue lift of 3.5% to 13.5%.

Does the Triple Whale study say AppLovin has higher ROAS than Meta?

Not directly. The study compares AppLovin with the other platforms in the study as a group and does not break the 2.08 benchmark out by channel. It shows AppLovin outperformed the brands' existing channel mix, which for most DTC brands is built around Meta.

Is AppLovin ROAS incremental or just attribution credit?

Triple Whale checked both. Attribution showed 2.90 vs 2.08 ROAS, and its media mix model and geo holdout tests also pointed to incremental revenue for most brands. Two of seven geo tests were not statistically significant, so results vary by brand, budget and test design.

How much should I spend to test AppLovin?

Triple Whale advises funding AppLovin like a real channel rather than a short 60 to 90 day trial, and results in the study were strongest for brands that gave it real budget and runway. The right number depends on your AOV and margins; our AppLovin ads cost guide covers typical starting budgets.

When was the Triple Whale AppLovin report published?

Triple Whale published "Does AppLovin Actually Drive Revenue? Here's What the Data Says" on September 3, 2026, written by co-founder Maxx Blank.