Published: 24th August 2026
A campaign structure that mixes every product, keyword and match type into one ad group does not save time. It hides waste, corrupts performance data and makes profitable scaling harder than it needs to be. Knowing how to structure Amazon campaigns is not about creating more campaigns for the sake of it. It is about giving each pound of spend a clear job: research demand, capture intent, defend margin or build category reach.
For growth-minded Amazon brands, the real objective is control. You need to see which search terms create profitable new-customer sales, which products are taking budget without converting, and where increasing bids will genuinely produce incremental revenue. That visibility starts with architecture.
Before building anything, establish the commercial constraints your advertising has to work within. ACoS is useful, but it is not a strategy on its own. A 25% ACoS may be excellent for one SKU and loss-making for another depending on contribution margin, VAT, fulfilment fees, discounts, returns and repeat purchase behaviour.
Set a target ACoS or, preferably, a target contribution margin after advertising for each product group. Then classify your catalogue. Products usually sit in one of four positions: proven profit drivers, growth products with conversion potential, launches that need controlled investment, and products that should not receive meaningful paid support until their retail fundamentals improve.
Do not ask advertising to rescue a weak listing, an uncompetitive price or a product with poor reviews. Fix the offer first. Traffic amplifies what is already there.
Campaigns should reflect how you intend to manage products commercially. Similar products can share a campaign only when they have comparable margins, prices, conversion rates and strategic roles. A £15 accessory and a £60 premium product may target similar terms, but they should not compete for the same budget or be judged against the same ACoS threshold.
For a focused catalogue, structuring campaigns at ASIN level provides maximum control. For a larger range, group close variants or a tightly related product family where performance is genuinely comparable. The test is simple: if one product needs a different bid, budget or profitability target, it needs its own line of control.
The strongest account structures separate discovery from performance. Research campaigns are allowed to explore. Performance campaigns are expected to convert. When the two are combined, expensive search-term testing quietly consumes the budget intended for proven revenue.
Use Sponsored Products automatic campaigns and broad or phrase match manual campaigns to find search behaviour you have not yet mapped. Automatic targeting is not a set-and-forget shortcut. It is a research tool, and it needs a defined budget, sensible bids and regular search-term analysis.
Where volume justifies it, separate automatic targeting into Amazon’s targeting groups: close match, loose match, substitutes and complements. This shows whether spend is coming from high-intent relevance or from broader product discovery. If the account is smaller, one automatic campaign can be enough initially, provided it is reviewed closely.
Broad match campaigns help uncover variations, use cases and longer search queries. Phrase match provides more control while still allowing useful discovery. Keep bids lower than your exact match campaigns because the uncertainty is higher. Your aim is not to make every broad target profitable immediately. Your aim is to identify terms worth graduating.
Once a search term has generated enough orders at an acceptable cost, move it into an exact match campaign. This is where you place stronger bids and give proven intent a protected budget.
Exact campaigns should be organised around tightly related keyword themes or individual high-value terms. A campaign containing ten unrelated keywords may still generate sales, but it gives you limited ability to manage bids, placements and budget allocation. Group terms that deserve the same commercial decision. Separate the terms that do not.
Add the promoted exact keyword as a negative exact match in the discovery campaign once it moves across. This prevents the same query competing against itself and keeps reporting cleaner. It also means your performance campaign receives the credit and budget it deserves.
This graduation process is the operational core of a mature Amazon account: discover, validate, isolate, scale, and keep the data clean.
Branded search behaves differently from category search. It normally converts at a stronger rate, carries lower customer acquisition cost and protects demand your brand has already created elsewhere. Non-branded terms are more competitive and often more expensive, but they are essential for acquiring customers who do not know you yet.
Keep branded Sponsored Products and Sponsored Brands campaigns separate from non-branded activity. Otherwise, low-cost branded conversions can make the account-level ACoS look healthier while non-branded acquisition quietly deteriorates.
Product targeting deserves its own structure too. ASIN and category targeting can be highly effective for stealing share from weaker competitors, defending your own detail pages and cross-selling complementary products. But it requires a different optimisation lens from keyword targeting. Review target-level conversion, competitor price position, ratings, delivery promise and relevance before increasing bids.
For larger brands, a practical structure often includes separate campaigns for competitor conquesting, own-ASIN defence and category discovery. For smaller budgets, prioritise the product-targeting role with the clearest commercial upside rather than spreading spend thinly across all three.
Sponsored Products should usually carry the bulk of conversion-led spend because they capture active shopping intent close to purchase. Sponsored Brands and Sponsored Display are not replacements. They extend the architecture when used with purpose.
Use Sponsored Brands for high-value category terms, brand defence and product-range storytelling. Store Spotlight can work well where a brand has distinct subcategories. Video often earns stronger engagement on mobile search results, but it still needs a clear proposition in the first seconds and a product detail page capable of converting the traffic.
Use Sponsored Display selectively for remarketing, detail-page defence and competitor audiences. It can support consideration and repeat exposure, but it is easier to waste money if audiences are too broad or attribution is assessed without context. Give it a controlled test budget and judge it against its role, not solely against last-click Sponsored Products efficiency.
Naming conventions sound administrative until a team cannot tell whether a campaign is research, brand defence or exact-match scale. Use a consistent format that states marketplace, ad type, product group, targeting type, match type and objective. Anyone reviewing the account should understand the campaign’s purpose without opening it.
Budget allocation should follow proven opportunity, not habit. Protect budgets for top-performing exact terms and priority product targets first. Then allocate a fixed testing allowance to discovery and launches. If a campaign regularly runs out of budget while delivering profitable sales, that is a scaling decision. If it runs out of budget while generating unprofitable clicks, it is an optimisation problem.
Negative keywords are equally strategic. Use negative exact matches to route graduated terms into their correct campaign. Use negative phrase matches to stop clearly irrelevant traffic. Avoid adding negatives too aggressively when search data is limited, as you can accidentally block valuable long-tail demand.
A well-structured account makes optimisation faster because the data is interpretable. Review budgets and major anomalies daily during high-volume periods. Review search terms, bids, placements and negatives weekly. Make larger structural decisions monthly, after allowing enough time for conversion data to settle.
Do not react to every single sale or non-sale. Amazon conversion data is noisy, especially with low-volume products. Establish minimum click and spend thresholds based on your average conversion rate and target ACoS before cutting a target. Equally, do not wait indefinitely for a weak keyword to prove itself. The point is disciplined decision-making, not false patience.
Placement adjustments should follow evidence. Top of Search may justify a premium for strong exact terms with reliable conversion. It is rarely sensible to apply blanket placement multipliers across an entire account. The same applies to dynamic bidding: use it where the campaign’s performance and margin can support it, not because it sounds more advanced.
The account you build should make every optimisation question easier to answer: what is this campaign meant to achieve, is it achieving it, and does it deserve more of the budget? If that answer is unclear, restructure before spending more. Profitable Amazon growth rarely comes from bidding harder. It comes from directing spend with intent.