Amazon PPC Campaign Structure (2026): Auto vs Manual, Naming Convention, Portfolios and a Launch Template

Campaign structure is the part of Amazon advertising nobody finds exciting and everybody pays for. A clean structure makes every later decision easy: the search term report maps to one ASIN, a high ACoS points to one campaign, a budget change affects one job. A messy one, which is what most accounts grow into, makes every report a blend and every fix a guess. This guide gives the structure we use for Indian sellers: what the auto campaign is for, how to split manual campaigns by match type, where brand defence, competitor and product targeting sit, how portfolios and budgets fit, a naming convention, the routine for migrating winning terms, and a launch template. The tools themselves are explained in our Amazon ads guide for Indian sellers.

Amazon's targeting guide gives one structural rule, which we follow throughout: create separate campaigns or ad groups for different objectives and "avoid mixing different targeting strategies within the same campaign and ad group". Everything else here is practice, not policy. Checked 25 September 2026.

What is the auto campaign for?

Research, not revenue. An automatic Sponsored Products campaign lets Amazon match your ad to searches and product pages using your listing content, across four match groups: close match (search terms closely related to your product), loose match (broadly related searches), substitutes (detail pages of similar products) and complements (detail pages of related products). Each can be bid separately or paused. Its output is the search term report, which tells you what shoppers actually type when they buy your product. Give it 10% to 20% of the budget, one ASIN or variation family per ad group, down-only bidding, and the seed negative list from our negative keywords guide. It should never be your only campaign after the first month, and it should never be paused, because the search language of your category changes.

How should manual campaigns be split by match type?

Amazon's three match types do different jobs, so they get different campaigns with different bids and budgets:

The reason to separate them is bid control. The same keyword in exact match converts better than in broad, so it deserves a higher bid; in one campaign you cannot give it one. And when a term is promoted to exact, it is added as a negative exact in the phrase and broad campaigns, so each search is served by exactly one bid and the reports stay clean. Sellers with under ten ASINs and small budgets can skip broad and run auto, exact and phrase.

Where do brand defence, competitor and product targeting go?

• Brand defence. One Sponsored Products exact campaign on your brand name and brand-plus-product terms, fixed bids, high top-of-search adjustment, small budget, and your brand name negated in every other campaign. Add a Sponsored Brands campaign on the same terms once you are in Brand Registry. Its ACoS will be very low; do not let that flatter the account average, and track it separately.
• Competitor keywords. A separate exact campaign on competitor brand names you have tested in auto and that convert. Expect ACoS 1.5x to 2x your generic campaigns. Keep it small and judge it on new-to-brand customers, not on ACoS alone.
• Product targeting. A manual campaign targeting specific ASINs (competitors with worse reviews or higher prices than yours, and your own complementary products for cross-sell) plus category targeting refined by price band, star rating and Prime eligibility. Placement is product pages, so keep top-of-search adjustment at 0% and judge it against the product-page row of the placement report.

How do portfolios and budgets fit?

A portfolio is a folder of campaigns with optional shared budget cap and date range. Use one portfolio per product line (or per brand if you sell several), and put every campaign for that line inside it. That gives you spend and ACoS per product line in one view, and a portfolio budget cap stops a launch line from eating the profitable line's money. Do not use portfolios by campaign type ("all exact campaigns"); the question you will ask later is "how is the kurta line doing", not "how are exact campaigns doing".

Set campaign daily budgets from the shares in the table above, then adjust fortnightly on "average time in budget": a profitable campaign that runs out more than 20% of the time gets more, an unprofitable one that never runs out gets lower bids. Remember Amazon can spend up to double the average daily budget on one day; the mechanics are in our budget rules and bidding guide. Ad spend sits on top of Amazon's fees, so set the account's total from margin using our Amazon seller fees guide.

What naming convention should you use?

The test is whether a stranger, or you in eight months, can read the campaign list and know what each one does without opening it. Our convention:

[Ad type] | [Product line] | [ASIN or family] | [Targeting] | [Job] | [Launch month]

Ad group names carry the ASIN or family and nothing else. Keyword-level notes go in a sheet, not in names. Once the convention exists, the search term report, the placement report and the bulk-operations file all sort into product lines by themselves.

How do winning terms migrate?

This is the loop that turns structure into falling ACoS. Every two weeks, on a 60-day search term report:

1. In auto, phrase and broad campaigns, find search terms with at least two orders and ACoS below target. These are winners.
2. Add each winner to the exact campaign for that ASIN at a bid slightly above what it earned in the source campaign.
3. Add the same term as a negative exact in the campaign it came from, so the source campaign keeps hunting for new terms and does not compete with the exact campaign in the auction.
4. Find search terms with zero orders above the click threshold and add them as negatives in the source campaign.
5. Find keywords in the exact campaign with 20 or more clicks and ACoS above target and cut their bids 15% to 20%. Do not negate a converter.
6. Where the search term report shows a competitor ASIN converting in auto (substitutes or complements), add it to the product-targeting campaign and negate it in auto.

Log each move with the date. In three months the exact campaign holds 30 to 80 proven terms per ASIN and carries most of the spend, the auto campaign has a long negative list and a small budget, and the account's ACoS is set by the exact campaign's bids, which you control. The full bid-and-negative routine is in our guide to reducing ACoS.

Launch template for one ASIN

For a new ASIN in a competitive category with a ₹1,500 daily budget, the structure on day one:

Add brand defence when brand searches appear in the report and Sponsored Brands when Brand Registry is approved. Expect the launch ASIN to run above break-even ACoS for six to eight weeks while reviews accumulate; the aim in that window is rank and review velocity, tracked on TACoS and organic position, not on the ACoS of the exact campaign. If after twelve weeks the exact campaign is still above break-even on its top ten terms, the listing or price is the problem, not the structure.

What does a badly structured account look like?

• One campaign per ad type with every ASIN in one ad group; the report cannot say which product wastes money.
• Auto, broad and exact targets in one campaign sharing one budget and one bidding strategy.
• Campaign names like "Campaign 3 copy" and "Test new".
• The same keyword live in exact, phrase and broad with no negatives between them, so you bid against yourself.
• Brand terms mixed into generic campaigns, so the account's ACoS looks fine while generic terms lose money.
• No portfolios, so a launch line's overspend is invisible until the monthly invoice.

Restructuring an account like this is the first job on most engagements our marketplace ads management team takes on, and it is done without pausing the old campaigns until the new ones have four weeks of data, so rank is not lost in the changeover.

FAQ

How many campaigns does one ASIN need? Four to start: auto, exact, phrase and product targeting. Add broad, brand defence and Sponsored Brands as the data and Brand Registry allow. Small catalogues can skip broad.

Should I put multiple ASINs in one ad group? Only tight variation families that share search terms, such as sizes of one kurta. Different products in one ad group blur the search term report and the bid decisions that follow.

Can I pause the auto campaign once exact is working? Keep it running on a small budget. It continues to surface new search terms and competitor ASINs as the category changes; pausing it stops the research.

What is the point of portfolios? Grouping campaigns by product line for reporting, and an optional shared budget cap so one line cannot overspend into another. They do not affect delivery.

How long before restructuring shows in ACoS? Four to six weeks, allowing for the up-to-14-day attribution window and time for the exact campaign to gather proven terms. Do not judge the new structure in week one.

Is a separate campaign per match type overkill for a small seller? Not if you want to control bids per match type, which is where the savings come from. A seller with three ASINs runs nine to twelve campaigns; that is manageable.

Building a structure you can read

Structure is the foundation for every bid, negative and budget decision that follows. If you would rather an Amazon Ads Advanced Partner built and ran it, starting with a free audit of the campaigns you have, see our marketplace ads management service.