This account was running close to break-even on ad spend. The brief was margin, not volume. Comparing Jan–May 2024 with Jan–May 2025, spend fell 36% and cost per acquisition 32% while return on ad spend held steady. Purchase volume fell as well — that is stated in full below, and both screenshots are published on this page.
↓36%
Ad Spend
↓32%
Cost per Acquisition
Held
Return on Ad Spend
↓39%
Purchases
The Situation
Break-Even Spending, at Scale
Between January and May 2024 the account spent $401K and produced 1.88K purchases at $213 each, at an actual ROAS of 105.35%. In other words, roughly a dollar back for a dollar spent — before the cost of the goods, the fulfilment or the team.
An account at 105% ROAS is not failing loudly. It is running a large volume of activity for very little margin, which is a comfortable place to sit for a long time. The brief was to stop doing that: keep the return, remove the spend that was not earning it.
Before — Jan–May 2024
$401K
Ad Spend
$213
Cost / Purchase
1.88K
Purchases
105.35%
Actual ROAS
After — Jan–May 2025
$258K
Ad Spend
$145
Cost / Purchase
1.14K
Purchases
102.68%
Actual ROAS
Spending heavily at close to break-even?
A free audit shows which of that spend is actually earning.
The Approach
One Decision. Margin Over Volume.
This was not a growth engagement and the work reflects that. The account was not restructured for scale; budget was withdrawn from the traffic that was not paying for itself.
1
Budget Pulled Out of Low-Intent Traffic
Spend was withdrawn from the segments producing clicks and impressions without producing purchases at an acceptable cost. In an account running at break-even, this is where most of the money is.
Low-intent segments cutWaste removed
2
Spend Concentrated on High-Intent Buyers
The remaining budget was concentrated on the audiences and queries with genuine purchase intent, accepting a smaller addressable pool in exchange for a materially better cost per acquisition.
High-intent focusNarrower, better pool
3
Campaign Quality Prioritised Over Volume
Reporting and optimisation targets were moved off click and impression volume. Once volume stops being the objective, the account stops defending spend that only produces volume.
Quality over volumeObjective reset
Results
36% Less Spend. 32% Lower CPA. Same Return.
Across the same five months a year later the account spent $258K, produced 1.14K purchases at $145 each, and recorded an actual ROAS of 102.68% — within three points of where it started.
Metric
Before
After
Change
Ad spend
$401K
$258K
↑ ↓36%
Cost per acquisition
$213
$145
↑ ↓32%
Actual ROAS
105.35%
102.68%
→ −2.5%
Purchases
1.88K
1.14K
↓ −39%
Revenue (derived)
≈$422K
≈$265K
↓ −37%
Revenue is derived — ad spend × actual ROAS, from the two figures shown in each screenshot.
Core Insight
$143K of spend was removed at effectively the same return. Whether that is a good outcome depends entirely on what the business wanted the channel to do — which is why the volume figures are in the table rather than left out of it.
The Evidence
The Screenshots Behind Every Number
Both panels below are the account's own Google Ads reporting for the two periods compared above. Every number in this case study, including the ones that fell, is read off them.
Jan–May 2024 — baselineGoogle Ads · 1 Jan – 27 May 2024
This is an efficiency case, not a growth case. Purchases fell from 1.88K to 1.14K and revenue with them, from roughly $422K to $265K. If the objective is top-line growth, these are the wrong numbers and we would not present them as anything else. If the objective is to stop spending $143K a year to stand still at break-even, they are the right ones. We publish it as the latter and leave you to judge which situation is yours.
Is This Your Account?
This case study is relevant if any of the following sounds familiar:
Your reported ROAS hovers near 100% and has done for several quarters
Spend has grown steadily year over year without margin growing with it
Nobody can say confidently which campaigns are earning and which are being carried
Volume targets are set before profitability targets, and the two have never been reconciled
You would trade a share of revenue for a materially better cost per acquisition
30 min · Free · We'll show you the gaps, no strings attached
Frequently Asked
Questions About This Case Study
Because it is what happened, and because the outcome was the intended one. This engagement had a margin objective rather than a growth objective. Presenting it as a growth case would require hiding the purchase and revenue figures — which are in the results table and visible in the screenshots. A case study you can check is worth more than one that only flatters.
It means the account returned roughly $1.05 of tracked revenue for every $1.00 of ad spend, before cost of goods, fulfilment, payment processing and overhead. For most physical-product businesses that is at or below break-even once real costs are applied. It is a number that looks acceptable in a platform report and is usually not acceptable in a P&L.
When the marginal spend is not earning its cost — typically visible as a long tail of campaigns or audiences sitting well below the account average, propped up by a handful of strong performers. If removing that tail leaves blended ROAS roughly unchanged, as it did here, the spend was not contributing. If blended ROAS falls when you cut, the tail was doing more than it appeared to and the diagnosis was wrong.
By segmenting the account by intent signal — query type, audience source, placement and device — and looking at cost per acquisition within each segment rather than the blended average. Segments that convert at several times the account CPA, or convert at volume with no margin, are the candidates. The check is whether removing them changes blended performance; if it does not, they were spending without earning.
Not as described, no. This account was deliberately made smaller and more efficient. A growth-stage brand needs the opposite sequence: find the segments that scale profitably first, then fund them. The technique that carries across is the diagnostic — segmenting by intent and reading cost per acquisition per segment. What you do with the answer depends on whether your constraint is margin or growth.
