AppLovin Case Study: Three Ecommerce Accounts, the Real Numbers, and How to Read Them

This AppLovin case study covers three anonymized WKND ecommerce accounts. A jewelry brand went from $268/day to $24.8K/day in 18 days at 2.94x NC D28 ROAS. A food and beverage CPG brand went from $111/day to $12.9K/day in 18 days at 2.52x. A footwear brand hit 3.52x NC D28 ROAS in an 18-day validation on $4,062.79 of spend. Those are AppLovin Ads platform metrics for new-customer cohorts, not blended MER or incrementality results. Below is what each number means, what it does not prove, what other brands have published, and a checklist for reading any AppLovin case study before you put budget behind it.

The short version

Account	Window	Spend ramp	Total spend	NC D28 ROAS	Other published metrics	What it is
Jewelry brand	March 4 to 21, 2026	$268.12 → $24,784.28 per day	$127,000	2.94x	$1.11 CPC	A scale case
Food & beverage CPG brand	January 24 to February 10, 2026	$111.12 → $12,954.25 per day	$68,437.26	2.52x	4.77% CTR	A scale case
Footwear brand	August 14 to 31, 2026	Not a ramp	$4,062.79	3.52x	$1.75 CPC · 4.78% CTR	“An efficient validation result, not a scale case”

• All three are anonymized and use AppLovin Ads new-customer cohort ROAS from the stated windows.
• WKND’s own caveat applies to all of them: “Platform attribution can differ from blended or incrementality measurement, and results vary by account.”
• None of the three has a published blended MER or incrementality result. That is the biggest gap to keep in mind.
• Public case studies from AppLovin and others (HexClad, MAËLYS, IPSY, Wayfair and more) are useful, but most are published by the platform itself or by agencies selling the channel. Read them with the checklist further down.

First, what “NC D28 ROAS” means

Every WKND number below is labeled NC D28 ROAS. Read that label carefully, because it controls what the number can and cannot tell you.

• ROAS is attributed revenue divided by ad spend, as AppLovin Ads reports it.
• NC (new customer) means the figure counts a new-customer cohort, not every buyer the platform touched. WKND’s services page describes these as “AppLovin Ads new-customer cohort metrics from the stated reporting windows.”
• D28 is one of AppLovin’s reporting windows. Metrics can be read on a D0 or D7 basis, and for many metrics longer windows such as D14 and D28 show the longer-term impact of ad clicks (AppLovin measurement and attribution).

New-customer cohort ROAS is a stricter platform view than all-customer ROAS, because returning buyers who would have bought anyway are not counted. It is still the platform’s own attribution. It says what AppLovin credited itself with, not what AppLovin caused.

That difference is the whole subject of our guide to platform ROAS versus incrementality. The short rule from WKND’s measurement framework: “no channel earns scaled budget on its own reporting, Axon included.”

AppLovin case study 1: jewelry brand, $268/day to $24.8K/day in 18 days

The published numbers (March 4 to 21, 2026):

Metric	Value
Daily spend, March 4	$268.12
Daily spend, March 21	$24,784.28
Total spend in the window	$127,000
NC D28 ROAS	2.94x
CPC	$1.11
Window	18 days

What the numbers mean

The spend ramp. Daily spend went from $268.12 to $24,784.28 in 18 days. By our arithmetic that is roughly 92 times the starting daily spend. Averaged across the window, $127,000 over 18 days works out to about $7,056 a day. So most of the money went out in the later days, which is what a ramp looks like.

The efficiency held while spend grew. A 2.94x new-customer ROAS on $127,000 means the platform kept finding new buyers at a price it reported as efficient while the budget climbed.

The CPC. At $1.11 per click, $127,000 bought roughly 114,000 clicks (again, our arithmetic). CPC tells you the traffic was cheap enough to test landing pages and offers at volume. It does not tell you conversion rate or order value, and those decide whether cheap clicks turn into profit.

What it does not show

• No blended MER or incrementality result is published for this account. We don’t know from the published case how total store revenue moved, or whether a holdout was run.
• It doesn’t say whether 2.94x was profitable. Profitability depends on the brand’s gross margin and CAC target. WKND’s framework puts breakeven ROAS at “1 divided by gross margin.” A 2.94x can be excellent for one margin structure and thin for another.
• It doesn’t show what happened after March 21. 18 days is a ramp, not a steady state. Whether the account held that level, plateaued or pulled back is not in the published case. If you’re past that stage and stuck, see why AppLovin spend stops scaling.
• It doesn’t isolate what caused the ramp. WKND’s services page describes the work generally: account and feed setup, ROAS-goal bidding, creative built for app placements, and measurement. The case doesn’t break results down by which of those mattered most, and we won’t invent that breakdown here.

For the full paid media program we run for jewelry brands, see our jewelry marketing agency page.

AppLovin ecommerce case study 2: food & beverage CPG brand, $111/day to $12.9K/day in 18 days

The published numbers (January 24 to February 10, 2026):

Metric	Value
Daily spend, January 24	$111.12
Daily spend, February 10	$12,954.25
Total spend in the window	$68,437.26
CTR	4.77%
NC D28 ROAS	2.52x
Window	18 days

What the numbers mean

A consumable product ramped on AppLovin. Food and beverage CPG is a different buying pattern from jewelry: lower price points and repeat purchase potential. This account shows that category can ramp on AppLovin Ads too. Daily spend went from $111.12 to $12,954.25, roughly 117 times the start by our arithmetic, and the window averaged about $3,802 a day.

The CTR. 4.77% of people who saw the ad clicked it. In app inventory, where the viewer is between game moments, CTR is mostly a creative signal: the hook, the pacing and the end card. WKND’s services page treats creative for app placements as its own discipline (“a format problem before it is a messaging problem”), and a high CTR is the first sign the format was working.

The ROAS. 2.52x NC D28 ROAS is lower than the jewelry account’s 2.94x. That comparison means little on its own. The two brands have different margins, prices and repeat rates, so their breakeven points differ. A CPG brand with strong repeat purchase may value a new customer beyond the 28-day window. Whether 2.52x cleared this brand’s target is not in the published case.

What it does not show

• No blended MER or incrementality result is published for this account.
• No CPC is published, so we can’t estimate click volume or cost per visitor the way we can for the jewelry brand.
• It doesn’t show retail or Amazon halo. CPG brands often sell in stores or marketplaces too. Several AppLovin-published cases (below) measured that halo. This one doesn’t report it.
• It doesn’t show lifetime value. D28 is 28 days. For a repeat-purchase product, the real value of these customers sits beyond that window.

AppLovin results from a validation, not a scale case: footwear brand

The published numbers (August 14 to 31, 2026):

Metric	Value
Total spend	$4,062.79
NC D28 ROAS	3.52x
CPC	$1.75
CTR	4.78%
Window	18 days

WKND’s services page labels it plainly: “This is an efficient validation result, not a scale case.”

What the numbers mean

This is the most useful case for a brand that hasn’t started. Most buyers don’t begin at $24K a day. They begin with a test and ask whether the channel deserves more. The footwear account answers the first half of that question: on $4,062.79, the platform reported 3.52x new-customer ROAS, the highest of the three accounts.

The traffic metrics. A $1.75 CPC and 4.78% CTR mean people were clicking at a healthy rate and the clicks cost more than the jewelry account’s. By our arithmetic, $4,062.79 at $1.75 is roughly 2,300 clicks. That is enough to see a pattern. It is not enough to be confident about how ROAS holds as spend grows.

If you’re planning a validation of your own, our guide to what a real AppLovin test budget looks like covers the platform minimum versus a test that can actually teach you something.

What it does not show

• It does not show scale. High ROAS at low spend often falls as budget rises, because the engine has to reach less responsive buyers. WKND’s own label says so.
• No blended MER or incrementality result is published. At this spend level, a blended read would almost certainly show nothing. WKND’s rule of thumb: “a channel under roughly 10% of total spend will not move blended MER detectably in a month.”
• It doesn’t tell you the next step. A validation earns a structured scale test, not a leap to full budget.

The three accounts side by side

Question	Jewelry	Food & beverage CPG	Footwear
Is it a scale case?	Yes, 18-day ramp	Yes, 18-day ramp	No, a validation
Platform metric	2.94x NC D28 ROAS	2.52x NC D28 ROAS	3.52x NC D28 ROAS
Blended MER published?	No	No	No
Incrementality published?	No	No	No
Anonymized?	Yes	Yes	Yes
Published by	WKND, on its services page	WKND	WKND

Two things are consistent across all three. The windows are short (18 days each), and every result is the platform’s own attribution. That makes them honest evidence that AppLovin Ads can deliver and ramp new-customer spend for ecommerce brands. They are not evidence that AppLovin was incremental for these brands, because that wasn’t measured in anything we’ve published.

Public AppLovin case studies from other sources

WKND’s three accounts aren’t the only public evidence. Here is what else was visible on September 27, 2026, grouped by who published it. For which kinds of brands run on the platform at all, see who uses AppLovin.

AppLovin-reported results

AppLovin’s partner page lists these brand results. All are AppLovin-reported:

Brand	Result as AppLovin reports it
HexClad	$1M+ incremental revenue in 3 weeks
MAËLYS	$200K/day within a week
Comfrt	400K+ hoodies
PROOF	+25% sales
Shapermint	8-figure revenue

AppLovin’s case-study library also carries named cases, including Wayfair (“How Wayfair uses AppLovin to find incremental growth”), IPSY (“drove a double-digit incremental lift in new subscriptions”), Ridge (“beats iROAS goal by 128%”), The Woobles (“achieves iROAS above target”) and Kitsch (“scales AppLovin spend while growing new-customer ROAS”). MaryRuth’s (“beats incremental ROAS target by 40%”) and HexClad (“drives $1M+ incremental revenue in 3 weeks”) case pages also appeared in Google results for applovin incrementality, and AppLovin’s YouTube channel carries a Knix case video, “How Knix Scaled Customer Acquisition with AppLovin“.

AppLovin’s consumer-brands page adds platform-wide claims, also AppLovin-reported: “$100K+ in daily ad spend reached within days of launch” and “70% of web brands see higher customer LTV with AppLovin vs. other platforms” (per its footnote, a 2025 AppLovin analysis of 95 brands).

How to read them: these come from the platform that sells the ads. That doesn’t make them wrong. Several of them (Wayfair, IPSY, Ridge, The Woobles) describe incrementality or iROAS, which is a stronger claim than platform ROAS. It does mean they are selected successes, and you should read the method section before you read the headline.

Agency and podcast case studies

• OMG Commerce published a podcast episode titled “How One Brand Went from $0 to $80K/Day on AppLovin in One Week” (YouTube, November 20, 2025). The episode features Miranda Pettinger of 365 Holdings (Cuddle Clones), and per its description the brand scaled AppLovin from zero to $80,000/day in 7 days during Black Friday. Black Friday is the highest-demand week of the year for most gift brands, so read that ramp as a peak-season result.
• metaply: a Google result for applovin case study ecommerce is titled “AppLovin case study: DTC glass brand, $595K attributed“. We’re reporting the title, not the method.
• bluealpha: a Google result for applovin roas is titled “Cann: AppLovin’s 2-16x ROAS Tested, Zero Incremental Lift“. Again, title only. It’s the clearest public reminder that a high platform ROAS and zero incremental lift can sit in the same account.

We haven’t read the full methods behind the metaply and bluealpha pieces, so we’re reporting their titles, not their conclusions.

How to read any AppLovin case study

Use these six questions on every case you read, including ours.

#	Question	Why it matters	Where our three cases stand
1	Is the number platform, blended or incremental?	Platform ROAS is what AppLovin credits itself with. Blended MER is total revenue ÷ total ad spend. Incremental ROAS is what a holdout says AppLovin caused. They answer different questions.	Platform only
2	What’s the window?	A 3-week or 18-day ramp is not a steady state. Short windows flatter because fatigue and saturation haven’t set in. Black Friday windows flatter because demand is at its peak.	18 days each, dated
3	New customers or all customers?	All-customer ROAS includes buyers who would have come back anyway. New-customer ROAS is stricter.	New-customer cohort
4	At what spend level?	3.52x on $4K and 2.94x on $127K are different achievements. Efficiency usually falls as spend rises.	Published for each
5	Validation or scale?	A validation shows the channel can work. A scale case shows it kept working as budget grew. Neither shows the other.	Two scale, one validation, labeled
6	Who published it, and why?	Platforms publish their best advertisers. Agencies publish their best accounts. Tool vendors publish results their tool measured. None is neutral.	WKND published them to sell its service. Apply the same skepticism to us.

Then add one more check that isn’t about the case at all: does the brand in the case look like yours? Margin, price point, repeat rate, creative volume and existing paid scale all change what AppLovin can do. A $200K/day beauty brand’s result says little about a $2K/day furniture store.

For the measurement side (holdouts, geo tests and how to compare incremental ROAS to breakeven), see platform ROAS versus incrementality.

What a case study can’t do for you

A case study can tell you a result is possible. It can’t tell you it will happen in your account. The only proof that counts for your budget is a test in your own account, read against your own margin:

1. Set the target from your economics. WKND sets the ROAS goal “from contribution margin and CAC.” Breakeven ROAS is 1 divided by gross margin.
2. Validate before you scale. Treat the first weeks like the footwear account: prove efficiency on a contained budget.
3. Scale in steps. WKND’s measurement framework says to scale in steps and retest at 2x once incremental ROAS clears breakeven by 30% or more.
4. Design the incrementality read before the test starts. When the CFO asks whether any of it is incremental, a holdout planned in advance is the only answer that holds up. A case study, ours included, can’t answer it for your account.

How WKND handles this

WKND runs AppLovin Axon ads management for ecommerce brands end to end: “account and feed setup, ROAS-goal bidding, creative built for app placements, and measurement that holds up against blended MER.” The three cases above are WKND accounts, published anonymized on that page with the caveat that platform attribution can differ from blended or incrementality measurement, and that results vary by account.

Across the practice, WKND reports $5M+ in Axon ad spend managed, 50+ Axon accounts and 5,000+ creative sets deployed. It’s led by owner-operators who, since 2018, have been responsible for driving more than $1B in revenue for 8- and 9-figure ecommerce brands.

On the creative side, one client note from the services page: “Those new end cards are sick!! CTR literally doubled” (Jenny Trinh, Daniel Patrick). That quote is not tied to any of the three cases above.

The fit is specific. WKND is “built for brands with proven product-market fit and existing paid scale,” and for ecommerce brands prepared to invest at least $500 per day. That $500 figure is WKND’s qualification threshold, not an AppLovin platform minimum or a performance guarantee. AppLovin itself lets you start with as little as $10 per day.

One honest note on our main case-study page: full case studies at WKND are general, and no other AppLovin cases are listed there today. The three accounts above are the AppLovin evidence we’ve published, and each has its own page: jewelry, food and beverage and footwear.

For the service behind these accounts, see all WKND AppLovin solutions.

FAQ

What is an AppLovin case study?

An AppLovin case study is a published account result for a brand advertising on AppLovin Ads, usually showing spend, ROAS and a time window. They come from AppLovin itself, from agencies like WKND, from measurement vendors and from podcasts. Check whether the number is platform, blended or incremental before trusting it.

Is there an AppLovin ecommerce case study with real numbers?

Yes. WKND publishes three anonymized ecommerce accounts with dated windows: a jewelry brand ($268/day to $24.8K/day in 18 days, 2.94x NC D28 ROAS), a food and beverage CPG brand ($111/day to $12.9K/day in 18 days, 2.52x) and a footwear validation (3.52x on $4,062.79). AppLovin also publishes named cases such as HexClad, IPSY and Wayfair.

What results can you expect from AppLovin?

No one can promise a result, and WKND’s own caveat says results vary by account. What the public cases show is that ecommerce brands have ramped AppLovin spend quickly at reported efficiency. What they mostly don’t show is blended or incremental impact. Expect to prove efficiency in a validation first, then scale in steps.

Is a 2.5 ROAS good on AppLovin?

It depends on your margin. Breakeven ROAS is 1 divided by gross margin, so at a 55% gross margin breakeven is about 1.82x (WKND’s illustrative example). A 2.5x platform ROAS would clear that on paper. Platform ROAS isn’t incremental ROAS, though, so confirm with a holdout before scaling on it.

What does NC D28 ROAS mean?

It’s AppLovin Ads return on ad spend for a new-customer cohort, measured over a 28-day window. It excludes returning customers, which makes it stricter than all-customer ROAS. It is still the platform’s own attribution, not a blended or incrementality measure.

How fast can AppLovin scale?

In WKND’s published accounts, a jewelry brand went from $268.12 to $24,784.28 in daily spend over 18 days, and a CPG brand went from $111.12 to $12,954.25 over 18 days. AppLovin reports that MAËLYS reached $200K/day within a week. Speed depends on creative, feed, tracking and the ROAS goal. Many accounts plateau when the goal is set tighter than the engine can hit.

Are AppLovin case studies reliable?

They’re reliable as evidence of what’s possible and unreliable as a forecast. Most are published by someone with a reason to show a win: the platform, an agency or a tool vendor. Look for the window, the spend level, new versus all customers, and whether any incrementality test was run.

Get your account read before you commit more budget

If you’re weighing AppLovin against a case study, the useful next step is a read of your own account or setup.

GET AN AXON ACCOUNT REVIEW → GOAL HISTORY · FEED AND EVENT HEALTH · CREATIVE FATIGUE · INCREMENTALITY READINESS. WRITTEN FINDINGS INSIDE FIVE BUSINESS DAYS.

Get an Axon account review · See the full case studies · AppLovin Axon ads management

Reporting shown is anonymized. ROAS figures are AppLovin Ads new-customer cohort metrics from the stated reporting windows. Platform attribution can differ from blended or incrementality measurement, and results vary by account.