Growth Performance

Amazon PPC campaign structure: control beats chaos

Most Amazon accounts do not have a bidding problem. They have a structure problem. When your campaigns are organised well, every other decision gets easier. Here is a clean way to build Sponsored Products.

Key takeaways
  • Separate discovery from performance so exploration and profit do not fight in the same budget.
  • Harvest winning search terms into exact-match campaigns and negate them where they were found.
  • Read ACoS and ROAS at the campaign level, because a clean structure is what makes the numbers legible.

Ask ten sellers why their Amazon ads underperform and most will talk about bids. The bid is rarely the root cause. The root cause is that everything lives in one messy campaign, so you cannot see what is working, cannot control spend, and cannot make a clean decision. Structure comes first. Bidding is what you do after structure gives you visibility.

This is about Sponsored Products, the workhorse of Amazon India advertising. Get the skeleton right and the rest of your account gets calmer.

The two jobs your campaigns are doing

Every Sponsored Products account is really doing two different jobs, and mixing them is where accounts go wrong.

  • Discovery. Finding out which search terms customers actually use to reach products like yours. This is exploration. It costs money and is supposed to.
  • Performance. Bidding hard and profitably on the terms you already know convert. This is exploitation. It should hit your target return.

When both jobs share one budget, discovery spend makes your performance numbers look bad, and performance targets starve your discovery. Separate them and each becomes legible.

Organising by match type

Match type is the backbone. A clean starter uses three layers:

  1. Automatic campaigns. Amazon decides where to show you. Pure discovery. Cheap way to surface search terms you never thought of.
  2. Broad and phrase campaigns. You seed keywords from your keyword research and let Amazon match loosely. More controlled discovery.
  3. Exact-match campaigns. Your proven winners, one term matched tightly, bid with intent. This is your performance engine.

Traffic flows down the funnel. Auto and broad find terms. Exact captures the winners. Each layer has a different job, a different bid logic and a different expectation.

Branded versus non-branded

Always split your own brand terms into a separate campaign. Branded searches convert cheaply because the customer already wants you. If you leave them mixed in, they flatter your blended ACoS and hide whether your prospecting on non-branded terms is actually profitable. Separation is honesty.

By product tier

Group products by how much they matter. Your hero SKUs, the ones that carry revenue and margin, deserve their own campaigns and closer attention. The long tail can share tighter, lower-maintenance campaigns. Do not give a low-margin filler SKU the same budget attention as your best seller.

The single-keyword ad group debate

Performance teams argue about single-keyword ad groups, where each ad group holds exactly one keyword. The upside is total control. You set one bid, and the report shows exactly what that term did, with no blending. The downside is a lot more campaigns to manage.

My practical take. Use single-keyword structure for your exact-match winners, where control pays off. For discovery, tight groups of a few closely related keywords are efficient enough. Do not build a thousand single-keyword ad groups on day one and drown in admin. Earn the complexity as terms prove themselves.

Harvesting search terms

This is the engine that makes the whole structure improve over time. Your auto and broad campaigns generate a search term report, the real phrases customers typed. Each week you do two things:

  • Promote converting terms into exact-match performance campaigns with a controlled bid.
  • Add those same terms as negative keywords in the discovery campaign that found them.

Why negate them where they were found? So your discovery budget stops paying to re-learn what you already know and keeps spending on finding new terms. Winners graduate up. Discovery keeps exploring. The account compounds.

Negative keyword discipline

The other half of the search term report is the terms that spend and never convert. Negate them. Irrelevant matches, wrong-intent searches, competitor terms that do not pay back. Every one you leave in place is a slow leak in your budget. Weekly negation is unglamorous and it is where quiet accounts beat loud ones. It is also how you protect your CTR and CVR from being dragged down by traffic that was never going to buy.

Reading ACoS and ROAS by campaign

Here is the payoff for all that structure. Because discovery, performance and branded are separated, you can read the numbers honestly.

Campaign type Expect Judge on
Exact performance Lower ACoS Profit and scale room
Auto and broad Higher ACoS New terms found
Branded Very low ACoS Defence, not discovery

A high ACoS on discovery is not failure, it is tuition. A high ACoS on your exact performance campaign is a real problem. Read ROAS the same way, by campaign role, never as one blended number that hides everything. When someone reports a single account ACoS with pride, they usually do not understand their own account.

Budget allocation

Put the majority of spend behind proven exact-match performance, a steady working share behind discovery so the pipeline of new terms never dries up, and a small defensive share on branded. Review weekly and shift budget toward whatever hits your target with room to grow. Budget follows evidence, not hunches.

A clean structure to start with

Build this and you have a foundation you can scale. One auto campaign per product group for discovery. One broad campaign seeded from keyword research. Exact-match campaigns for proven winners, single-keyword for the top terms. A separate branded campaign for defence. A weekly ritual of harvesting winners and negating waste. Read ACoS and ROAS per campaign, never blended. Do this consistently and your account gets more efficient every week, which is exactly what disciplined Website Development and off-Amazon funnels cannot substitute for.

The daily brief

Never miss a move

The moves that move money, every morning.

One email a day. No spam, ever.

FAQ

Quick answers.

Single-keyword ad groups give you clean control over bid and reporting for each term, which is why many performance teams like them for exact match. The cost is more campaigns to manage. For a small catalog, tight ad groups of a few closely related keywords are fine. Reserve single-keyword structure for your proven winners.
Keep branded terms in their own campaign. Branded traffic converts cheaply and would otherwise flatter your blended ACoS and hide how your prospecting is really doing. Separating them lets you read defence and discovery honestly and set different targets for each.
Your automatic and broad campaigns discover the real phrases customers type. When a search term converts well, you promote it into an exact-match campaign with a controlled bid, then add it as a negative in the campaign that found it. That way discovery keeps exploring and performance keeps the winners.
Put the majority behind proven exact-match performance campaigns, a working share behind discovery to keep finding new terms, and a small defensive share on branded. Review weekly and move budget toward whatever is hitting your ACoS target with room to scale.
Make it a weekly habit. Review the search term report, negate irrelevant or wasteful terms, and promote the good ones. Negative keyword discipline is what stops your budget leaking into searches that never convert.

Where Zane fits

Related insights

From the wire

India's Commerce Engine

Put it
to work.

hello@zane.marketing

Book a meeting