An auto parts store had built its whole account around buying clicks as cheaply as possible. Between June and August 2024 we rebuilt it around profit instead — with ad spend essentially unchanged at roughly ₴19K a month. Both account screenshots are published further down this page.
+49%
Revenue
×20
Conversions
↓18.4×
Cost per Conversion
Flat
Ad Budget
The Situation
An Account Optimised for the Wrong Number
In June 2024 the account produced 18.24 conversions at ₴1,000 each against ₴66.1K of revenue, at an average CPC of ₴5.44. Structure, bidding and keyword selection were all pointed at a single goal: buy clicks as cheaply as possible.
Cheap clicks and cheap orders are not the same thing. At ₴1,000 per conversion on a parts catalogue, the account was paying premium prices for outcomes while congratulating itself on its cost per click.
Before — June 2024
18.24
Conversions
₴1,000
Cost / Conversion
₴66.1K
Revenue
₴5.44
Average CPC
After — August 2024
357.76
Conversions
₴54.48
Cost / Conversion
₴98.6K
Revenue
₴4.68
Average CPC
Optimising for cost per click instead of profit?
A free audit shows what your account is actually buying.
The Approach
Four Changes. All Pointed at Profit.
The work was structural rather than tactical. Nothing here is a bidding trick — it is a rebuild of what the account was being asked to optimise toward.
1
Campaigns Restructured Around Profitability
Every campaign was rebuilt around profit-based segmentation rather than click cost. Keywords and ads that consumed budget without producing orders were removed rather than bid down.
Performance Max was introduced for coverage across Search, Shopping and the wider inventory, paired with retargeting for abandoned carts and product views so that existing intent was not left to cool.
Bidding moved to automated strategies, with manual exclusions kept in place so the algorithm could not drift back toward the cheap, unprofitable inventory the old account had been feeding on.
Automated biddingManual exclusionsDrift control
4
Faster Pages and a Streamlined Checkout
Landing pages were sped up and checkout friction removed. Traffic quality improvements are wasted if the site loses the visitor between the product page and the order.
By August 2024 the account was producing 357.76 conversions at ₴54.48 each against ₴98.6K of revenue — with the budget effectively where it started.
Metric
Before
After
Change
Conversions
18.24
357.76
↑ ×19.6
Cost per conversion
₴1,000
₴54.48
↑ ↓18.4×
Revenue
₴66.1K
₴98.6K
↑ +49%
Average CPC
₴5.44
₴4.68
↑ ↓14%
ROAS (derived)
×3.62
×5.06
↑ ×1.40
Ad spend (derived)
≈₴18.2K
≈₴19.5K
↑ +7%
ROAS and ad spend are derived — spend is conversions × cost per conversion, ROAS is revenue ÷ spend, both from figures shown in the screenshots.
Core Insight
The budget did not meaningfully move. Revenue rose 49% because the same money was pointed at orders instead of clicks — which is the whole difference between a cheap account and a profitable one.
The Evidence
The Screenshots Behind Every Number
Both panels below are the account's own Google Ads reporting for the two months compared above. Every number in this case study is read off them.
Ad spend was not literally identical. Derived from the screenshots it rose from roughly ₴18.2K to ₴19.5K — about 7%. We describe the budget as flat because a 7% move over two months is small next to a 49% revenue gain, but the figures are here rather than rounded away.
Is This Your E-Commerce Account?
This case study is relevant if any of the following sounds familiar:
Your reporting leads with cost per click, and cost per order is somewhere further down the page
Campaigns are organised by whatever structure the account was set up with years ago, not by margin
Underperforming keywords get bid down rather than removed, and never quite go away
Abandoned carts and product views are not retargeted, so earned intent is left to cool
You suspect the account could do more on the same budget but cannot prove where the waste is
30 min · Free · We'll show you the gaps, no strings attached
Frequently Asked
Questions About This Case Study
It is arithmetic on the two screenshots: ₴1,000 in June against ₴54.48 in August. The size of the multiple comes from how bad the baseline was — the June account was recording very few conversions for the money it spent, so almost any structural improvement produces a dramatic ratio. The revenue figure is the more meaningful one: +49% on an unchanged budget.
Not exactly, and we say so on the page. Derived from the screenshots, spend went from roughly ₴18.2K to ₴19.5K — a 7% increase. That is a rounding-scale difference next to a 49% revenue gain, but it is a difference, and a case study that hid it would not be worth reading.
It means the account is segmented by what products actually earn, not by what is convenient to group. High-margin and high-volume lines get their own campaigns and their own budgets; low-margin lines are capped or excluded rather than left to absorb spend inside a mixed campaign. Bidding targets are then set per segment instead of one blended target across the catalogue.
It is effective if you constrain it. Left unconstrained, PMax will find the cheapest available inventory, which for a parts catalogue often means low-intent traffic on generic terms. The controls that matter are exclusions, a well-structured feed, and campaign-level separation so PMax cannot quietly absorb budget from the campaigns that were already working. That is why automated bidding here was paired with manual exclusions.
Partly because the restructure moved spend away from expensive generic terms toward more specific, lower-competition queries, and partly because improved landing pages and quality signals reduce what you pay for the same position. The CPC drop is modest — 14% — and is the smallest of the changes here. The conversion rate improvement did most of the work.
