Published: 18th August 2026
A healthy-looking ACoS can hide a weak Amazon advertising account. That was the issue in this Amazon campaign restructure case study: sales were growing, but the account had become a tangle of overlapping campaigns, inconsistent bidding and budgets that ran out before the best terms could scale.
The brand did not need more campaigns. It needed a structure that made every pound of spend accountable.
Over a 90-day period, a full campaign rebuild reduced wasted spend by 34%, improved advertising cost of sales from 31.8% to 24.6%, and lifted attributed sales by 27%. The improvement did not come from a blunt bid reduction. It came from deciding exactly where each search term, product target and budget belonged.
The figures have been rounded to protect commercially sensitive data, but the operating principles are directly applicable to established UK brands facing the same problem.
The account had been managed in a familiar pattern. New Sponsored Products campaigns were created whenever a product launched, a keyword was found, or a performance issue appeared. Automatic campaigns, broad match campaigns and exact match campaigns were all targeting many of the same high-volume terms. Product targeting was mixed with keyword targeting. Branded terms sat beside generic acquisition terms.
On the surface, this produced volume. Underneath, it created auction overlap.
The same search term could receive bids from three or four campaigns, each using a different bid level and optimisation logic. The brand could not tell which campaign deserved budget, which search terms were genuinely incremental, or whether a sale came from a high-intent shopper or from branded demand it would probably have captured anyway.
Budget pacing made the problem worse. Several campaigns spent their daily allowance by early afternoon. The highest-converting exact-match terms were constrained while broad campaigns continued to absorb spend on loosely relevant queries. The team was reviewing results, but it was optimising inside a structure that could not give a clear answer.
This is where many Amazon accounts stall. The issue is not effort. It is architecture.
Before changing bids, we mapped the account at search-term level. That matters because campaign-level metrics can be misleading. A campaign showing a 25% ACoS may contain one profitable search term, several break-even terms and a large number of clicks that will never convert.
The audit focused on five areas: search-term duplication, budget allocation, match-type control, product targeting, and the relationship between advertising performance and retail readiness.
The brand’s strongest generic terms had high conversion rates but limited visibility because they were trapped in mixed ad groups. At the same time, automatic campaigns had surfaced useful long-tail queries but were allowed to keep bidding on terms that had already been proven in manual exact campaigns.
There was also a commercial issue. A small group of hero ASINs drove most profitable revenue, yet spend was spread too evenly across the catalogue. Lower-margin products were receiving aggressive bids without a clear role in the wider growth plan.
No campaign restructure should start with a fixed template. The right structure depends on catalogue depth, margin, conversion rate, review strength, stock cover and the brand’s immediate objective. A product with thin margin may need a strict efficiency target. A strategic new launch may justify a controlled period of higher ACoS. Treating both in the same way produces bad decisions.
The rebuild separated discovery, control and defence. Each campaign type had one job, one budget logic and one route for moving search terms through the account.
Discovery campaigns used automatic targeting, broad match and selected product targets to find new demand. Their purpose was not to carry the account’s revenue target. Their purpose was to generate clean search-term data at an acceptable cost.
Control campaigns held proven, non-branded terms in phrase and exact match. Exact campaigns were reserved for terms with enough conversion evidence to warrant dedicated bids and protected budget. This gave the brand a reliable place to scale the traffic it wanted most.
Defence campaigns isolated branded search terms. That made branded ACoS visible rather than allowing it to flatter generic acquisition performance. Brand defence can be commercially sensible, particularly where competitors are active, but it should never obscure the cost of winning new customers.
Product targeting was rebuilt separately from keyword activity. Competitor ASINs, complementary products and defensive targets each received distinct campaign groupings. This allowed bids to reflect the quality of the target rather than being averaged across a mixed list of products.
Negative keywords were the control mechanism that held the structure together. Once a term moved from discovery into exact match, it was blocked in the relevant discovery campaigns. This reduced internal competition and ensured performance data remained interpretable.
The objective was not perfect isolation for its own sake. Excessive segmentation can create too many low-volume campaigns, slow optimisation and make budgets harder to manage. The objective was practical control: enough separation to make informed decisions without turning the account into an administrative burden.
Campaign structure without budget discipline simply moves the problem around.
The previous account allocated daily budgets based largely on historic campaign spend. The restructure allocated budget according to commercial role. Hero ASINs with strong conversion and stock cover received priority. Exact-match campaigns protecting proven generic terms were funded before discovery activity. Branded campaigns had caps that reflected their defensive purpose rather than their ability to generate cheap sales.
Budget was reviewed against time of day and stock position, not just a monthly target. If an exact campaign was budget-limited while meeting its efficiency threshold, it received incremental spend. If a discovery campaign was spending without producing viable search-term candidates, its budget was reduced or redirected.
This is a crucial distinction. Lowering a bid can improve ACoS while reducing profitable sales. Increasing a budget can worsen ACoS while increasing contribution profit. The correct decision depends on margin, repeat purchase behaviour, stock availability and the brand’s wider objective – not a single dashboard metric.
The first gains came from waste reduction. Search-term negatives removed repeated spend on irrelevant and duplicated traffic. Next came improved conversion efficiency as proven terms received dedicated bids and more consistent budget.
By day 90, attributed sales had increased by 27% while ad spend rose by only 9%. ACoS moved from 31.8% to 24.6%. More importantly, the account was no longer dependent on a handful of broad campaigns to produce revenue.
The share of spend going to exact-match non-branded campaigns increased, while the proportion consumed by poorly controlled discovery activity fell. Product targeting became easier to evaluate because competitor targets were no longer bundled with defensive targets. The team could see which ASINs deserved further investment and which were merely generating clicks.
Conversion rate also improved, but advertising was not given sole credit. Listing content, price positioning and stock availability remained part of the equation. Amazon PPC cannot compensate indefinitely for a weak detail page or an uncompetitive offer. A senior account strategy connects ad decisions to the retail fundamentals that determine whether paid traffic can convert.
A restructure is not a one-off clean-up. It is the foundation for a better operating rhythm.
Search-term reports were reviewed weekly, with clear rules for promotion, negation and bid adjustment. Budgets were monitored more frequently during peak trading periods and around stock changes. Performance was judged by campaign purpose, not by one universal ACoS target.
That last point matters. A branded defence campaign, a generic exact campaign and a product-targeting discovery campaign should not be expected to perform identically. They serve different roles in the funnel and should be managed against different thresholds.
The brand also stopped making account-wide changes based on short windows of data. Amazon performance is affected by seasonality, deal activity, competitor behaviour and conversion shifts. A disciplined testing approach protects against reactive optimisation that damages the terms already driving profitable demand.
A full rebuild is justified when campaign overlap prevents clear decisions, spend is repeatedly exhausted on low-value activity, or search-term reports show the same terms appearing across multiple campaigns without a reasoned strategy.
It is not always the right answer. If an account is small, data is limited and campaigns are already logically separated, a targeted clean-up may be more effective than a wholesale restructure. Rebuilding too aggressively can reset learning, interrupt sales momentum and create unnecessary operational risk.
For established brands, however, a controlled restructure is often the fastest route to clarity. It gives leadership a view of where Amazon advertising is creating incremental growth, where it is defending existing demand, and where budget is being wasted.
The real value is not simply a lower ACoS. It is an account that can be directed with confidence. When every campaign has a role, every search term has a home and every budget has a commercial purpose, profitable scale becomes a management decision rather than a hope.