How to Structure Profitable Amazon PPC Campaigns
Most struggling PPC accounts don't have a bidding problem — they have a structure problem. When keywords are jumbled together and auto and manual campaigns overlap, you can't tell what's working, and optimization becomes guesswork. A deliberate structure fixes that. Here's the framework that keeps campaigns clean, measurable, and scalable.
Try Helium 10 free — then save with our link
Start on the free plan and unlock member savings on Platinum & Diamond. Cancel anytime.
The hierarchy: account, campaign, ad group
Amazon ads nest in three layers. The account holds your campaigns; each campaign holds one or more ad groups; each ad group holds keywords or product targets and the ads themselves. Budgets are set at the campaign level, while bids live at the keyword and ad-group level. The mistake beginners make is stuffing unrelated keywords into a single ad group — that averages out performance and hides your winners and losers.
Structure is just the discipline of keeping things separate enough that the data tells you a clear story.
The auto-to-manual flow
The cleanest structure follows the natural lifecycle of a keyword: discover it, then scale it. Set it up as three linked campaigns per product.
- Discovery (automatic): a single auto campaign that lets Amazon match your listing to searches you haven't thought of.
- Research (manual broad/phrase): promising terms from discovery moved into broad or phrase match to explore variations.
- Performance (manual exact): proven converters isolated into exact match where you bid confidently and control ACoS.
As a keyword proves itself, it graduates up the chain, and you add it as a negative in the lower campaign so the same term isn't competing against itself.
One product theme per campaign
Group keywords by theme and match type, not by whim. A stainless steel water bottle might have separate ad groups for "insulated water bottle," "steel water bottle," and "gym water bottle." Tight themes let you write relevant ads and set bids that reflect how each cluster actually performs.
A simple starting blueprint
| Campaign | Match type | Job | Bid posture |
|---|---|---|---|
| Auto — Discovery | Automatic | Find new search terms | Low, let data accumulate |
| Manual — Research | Broad / phrase | Test keyword variations | Moderate |
| Manual — Performance | Exact | Scale proven winners | Higher, protect top spots |
| Manual — Brand defense | Exact | Own your branded terms | Efficient, low competition |
Quick tip
Use a consistent naming convention like SP | Product | Match | Purpose. When you have dozens of campaigns, a good name tells you instantly what each one does.
Single vs. multiple keywords per ad group
There's a long-running debate between the "single keyword per ad group" (SKAG) approach and grouping several related keywords together. SKAGs give you pinpoint control — each keyword gets its own bid and its performance is never diluted by neighbors — but they multiply the number of campaigns you have to manage. Grouping a handful of tightly related terms is more maintainable and still clean enough to read the data. For most sellers, small themed ad groups strike the right balance; reserve SKAGs for your highest-value head terms where precise bid control genuinely pays off.
Isolating your best sellers
If one product drives most of your revenue, give it its own dedicated campaign rather than lumping it with slower movers. Sharing a budget across products means your winner can get starved when a weaker product burns through the daily cap early in the day. Separate campaigns let you fund your hero product properly and protect its impression share during peak buying hours.
Setting budgets and bids
Give your performance campaigns the largest budgets since they carry proven keywords, and keep discovery budgets modest — they exist to learn, not to scale. Start bids near Amazon's suggested range, then adjust based on whether each keyword hits your target ACoS.
Keeping the structure alive with Adtomic
The auto-to-manual flow only works if you actually harvest keywords and add negatives regularly. Helium 10's Adtomic automates that maintenance: it flags high-performing search terms ready to graduate, suggests which terms to negate, and applies bid changes across your structure. Pair it with Cerebro to reverse-engineer competitor keywords before you build the campaigns in the first place, so your structure starts from real search demand.
Reviewing and pruning your structure
A structure isn't something you build once and forget. Over time campaigns accumulate cruft — ad groups that never gained traction, keywords that graduated but were never removed from their source, budgets that no longer match performance. Schedule a periodic review to consolidate underperformers, retire dead ad groups, and rebalance budgets toward the campaigns actually driving profit. A tidy account is easier to read, and an account you can read is one you can actually improve. Left unpruned, even a well-designed structure slowly turns back into the tangle it was meant to prevent.
The bottom line
A profitable PPC account is a well-organized one. Separate discovery from performance, keep each ad group thematically tight, and let keywords graduate as they prove themselves. Build the structure once and every optimization decision afterward becomes obvious.
Ready to put this into action?
Helium 10 gives you the tools in this guide in one dashboard. Start free through our link.