Amazon Campaign Structure: How to Build a PPC Architecture That Scales

Most Amazon advertising accounts grow by accumulation. New campaigns get layered on top of old ones reactively, ad groups balloon with mismatched keywords, match types blend together, and after eighteen months the account looks like a filing cabinet nobody has cleaned out since launch.

The data is there, but you cannot read it.

You cannot optimize what you cannot see, and you cannot scale what you cannot optimize. This is the wall most intermediate sellers hit between one million and five million in annual revenue.

The fix is not more spending or more tools. It is a deliberate campaign architecture that separates your data, isolates your variables, and gives you a clean signal to act on as you grow.

In this post you will learn the three structural layers every Amazon advertising account needs, the six-campaign framework that organizes Sponsored Products from the ground up, the ad group and match type rules that keep your data clean, and the budget allocation that turns the whole architecture into a scalable system.

Why Campaign Structure Is the Highest-Leverage PPC Decision You Will Make

When you run a manual campaign with broad, phrase, and exact match keywords all targeting the same ad group, you have created a data swamp. You cannot tell which match type drove the conversion. You cannot bid up an exact match without bidding up broad. You cannot expand your keyword set without polluting your top performers.

Campaign structure is the discipline of giving every objective, every keyword type, and every product set its own clean container so bid changes, budget shifts, and keyword decisions only affect the lever you are pulling. The structure is invisible to shoppers. It is everything to the operator behind the account.

The Three Layers of Account Architecture

Every Amazon advertising account has three structural layers that must be designed deliberately.

Campaign layer:

Your budget and bidding strategy container. Each campaign holds one budget, one bidding strategy, and one date range. The level at which you control ad spend.

Ad group layer:

Your targeting container. Each ad group holds one match type, one tightly themed keyword cluster or product target set, and a defined group of ASINs. The level at which you control relevance.

Keyword and target layer:

Your bid container. Each keyword or product target holds one bid, and that bid is the only lever you adjust at this level. The level at which you control unit economics.

When these three layers are confused, you cannot diagnose what is working. When they are clean, every report tells you exactly what to do next. For the full strategic context behind these decisions, our Amazon PPC management guide goes deeper on the philosophy underneath the architecture.

The Core Campaign Framework for Sponsored Products

Sponsored Products is where most accounts spend the majority of their ad budget, and where structural discipline pays off fastest. The framework below works for accounts from one million to fifty million in revenue, with adjustments at the edges.

1. Auto campaigns for discovery. 

Every product or tightly related product set gets an ongoing auto campaign for keyword and ASIN harvesting. Set bids modestly. The job is not volume, it is surfacing search terms and competitor ASINs you would not have thought to target manually. Mine the search term report weekly during launch and monthly after stabilization.

2. Manual exact campaigns for proven converters. 

Every keyword that proves itself in auto or in a research pass gets graduated into a manual exact campaign. Exact match gets its own campaign, not just its own ad group. This is the campaign you fund aggressively, where your ACoS targets are tightest, and where you protect your unit economics.

3. Manual phrase campaigns for expansion. 

Phrase match lives in its own campaign so you can expand your reach into related queries without competing against your own exact terms. Phrase is where you fish for new converters that will eventually graduate into exact.

4. Manual broad campaigns for discovery at scale. 

Broad match is high-spend, high-variance, and prone to wasted clicks if mixed with anything else. Give it its own campaign so you can cap budget independently and harvest aggressively from its search term report.

5. Product targeting campaigns for ASIN-level competition. 

ASIN targeting has fundamentally different bidding dynamics than keyword targeting. Conversion rates, click rates, and intent profiles all differ. Keep it separated by campaign so you can manage budget and bids without contaminating your keyword data.

6. Defensive campaigns for your own brand. 

A separate campaign targeting your own brand keywords and your own ASINs prevents competitors from stealing high-intent traffic and keeps your branded search ACoS extremely low. Non-negotiable for any brand serious about brand protection on Amazon.

Six campaign types per product cluster, not per individual ASIN. Group products by shared keyword universe, not by SKU count.

How to Structure Ad Groups Inside Each Campaign

The campaign tells the platform how much to spend. The ad group tells it what to spend that money on. Two principles govern good ad group structure.

One match type per ad group, always. Even within a manual exact campaign, do not mix exact and phrase variants. The cleaner your match type isolation, the more accurately you can diagnose what is converting.

One tight theme per ad group. A magnesium glycinate ad group should not contain magnesium citrate keywords. A women's running shoe ad group should not contain men's running shoe keywords. When themes blur, your conversion rates blur, and your bid decisions stop working. A useful rule: if you cannot describe the ad group's theme in one short phrase, it is too broad.

For brands with deep catalogs, this means more ad groups, not fewer. A larger number of tightly themed ad groups is easier to manage than a smaller number of bloated ones, because each ad group's report tells you exactly one story.

Match Type Strategy: Where the Most Common Mistakes Live

The most common structural mistake in intermediate accounts is treating match types as interchangeable. They are not. Each plays a distinct role, and they should never compete for the same query in the same auction.

Exact match is for queries you have already proven. You know the conversion rate. You know the ACoS at which it pays. You bid to your target unit economics and protect that position.

Phrase match is for queries you suspect will convert but have not yet validated. You bid lower than exact, watch the search term report, and harvest winners into exact matches.

Broad matches are for discovery. You bid lower still, accept higher ACoS, and treat the spend as a research budget. Anything that converts well in broad gets harvested into phrase or exact, where it will perform better at lower cost.

When the same keyword exists in all three match types simultaneously, use exact-match negation in your phrase and broad campaigns. This forces the query to flow to your exact campaign, where you have already optimized the bid. Without that negative structure, your phrase and broad campaigns steal traffic from your exact campaign at higher cost. This is the single highest-impact negative keyword move in most accounts.

Sponsored Brands and Sponsored Display Fit Into a Separate Layer

Sponsored Products is the foundation, but a complete architecture includes upper-funnel and remarketing layers.

Sponsored Brands campaigns should be organized by brand defense, category capture, and competitor conquest, each in its own campaign with distinct creative. Headline placement and the video formats convert differently from Sponsored Products, and the data needs to live separately to be readable.

Sponsored Display lives in its own structural tier. Product targeting, audience targeting, and view remarketing each serve different intent points in the funnel. Treating them as one campaign type collapses three different optimization conversations into one undifferentiated dataset.

Budget Allocation: What Sits Above the Structure

A useful starting allocation for a mature Sponsored Products account looks roughly like this: 50 to 60 percent of budget to manual exact, 15 to 20 percent to manual phrase, 10 percent to broad and auto for discovery, 10 to 15 percent to product targeting, and a small dedicated allocation to defensive campaigns. Sponsored Brands and Sponsored Display sit on their own line item, typically 15 to 30 percent of total ad spend depending on category maturity.

These ratios are a default that gets adjusted by category, competitive intensity, and where you are in the launch curve. The point is that with a clean structure, you can actually see these ratios and adjust them deliberately. Without a clean structure, you are guessing.

FAQ

How should I structure my Amazon PPC campaigns?

Build your account around three structural layers. The campaign layer controls budget and bidding strategy, the ad group layer controls targeting and relevance, and the keyword layer controls bids and unit economics. Within Sponsored Products, organize your campaigns by purpose using six core campaign types: auto, manual exact, manual phrase, manual broad, product targeting, and defensive. Each campaign should have one budget, one bidding strategy, and one clear job.

Should I separate match types into different campaigns or different ad groups?

Always separate match types at the campaign level, not just the ad group level. Exact, phrase, and broad each have different bidding logic, different ACoS targets, and different roles in your funnel. Keeping them in separate campaigns lets you control budget independently for each match type and prevents your discovery spend from cannibalizing your proven converters. This is the single most impactful structural decision in most intermediate accounts.

How many Amazon PPC campaigns should I have?

There is no universal number, but a useful baseline for a mature account is six core campaign types per product cluster, not per ASIN. Group your products by shared keyword universe rather than by SKU count, then build auto, manual exact, manual phrase, manual broad, product targeting, and defensive campaigns for each cluster. A brand with three distinct product clusters might run roughly eighteen Sponsored Products campaigns plus separate Sponsored Brands and Sponsored Display layers.

What is the best budget allocation for Amazon Sponsored Products?

A reasonable starting allocation for a mature Sponsored Products account is 50 to 60 percent of budget to manual exact campaigns, 15 to 20 percent to manual phrase, 10 percent to broad and auto for discovery, 10 to 15 percent to product targeting, and a small dedicated allocation to defensive campaigns. These ratios shift based on category, competitive intensity, and how mature your products are in their launch curve.

Why does my Amazon PPC ACoS keep climbing as I scale?

Climbing ACoS during scale is usually a structural problem, not a bid problem. When match types share campaigns, ad groups mix themes, and the same keyword competes against itself in three places, your spend gets cannibalized and your data becomes unreadable. The fix is structural separation: isolate match types into their own campaigns, tighten ad group themes, and apply exact-match negation in your phrase and broad campaigns to stop internal competition.

The Bottom Line

Campaign structure is invisible work. It does not look impressive in a screenshot. What it produces is a clean account where every report is readable, every bid change is meaningful, and every new product launch has a home that already exists. 

The brands that scale efficiently past five million on Amazon are almost universally the brands that took architecture seriously before they took growth seriously. If your account has accumulated rather than been designed, the structural rebuild is the highest-leverage advertising decision you can make this quarter. 

At Amazon Growth Lab, the first thing we look at on any new account is the campaign architecture, because everything else downstream depends on it.

See how this approach has worked for other brands, or book a free audit and we will show you exactly where your current structure is costing you.