Per-Product or Per-Keyword-Cluster? The Campaign Structure Question Nobody Explains

Before you've spent a single dollar on Amazon PPC, you'll hit a decision most guides skip straight past: should each product get its own campaign, or should you group products together under shared keyword themes? Get it wrong and you usually won't find out for months, right around the time you're staring at a reporting dashboard that doesn't match how your business actually works.

Why this decision is easy to get backwards

The instinct for a lot of new sellers is to mirror their product catalog: one campaign per product, because that's how the seller thinks about their own inventory. It feels organized. It's also exactly backwards for a big chunk of catalogs, because Amazon's ad platform doesn't care how you organize your warehouse. It cares whether two listings are actually competing for the same searches.

The real question: do these products share an audience?

Forget the product for a second and ask what someone searching Amazon is actually looking for. If two listings would show up for basically the same search term (two color variations of the same water bottle, say), splitting them into separate campaigns doesn't organize anything. It just splits your data in half. You'll spend twice as long learning the same thing, on half the budget each.

If two listings are for completely unrelated products, the opposite is true: cramming them into one campaign muddies your reporting, because a click or a sale can't be cleanly attributed to the product that actually earned it.

The four situations, and what actually fits each one

• Straight variations of the same listing (different size or color, same core product): keep these in one campaign, split into ad groups by match type instead of by variation. They share an audience, so don't fragment the data. See the variation-blending mistake for exactly how this goes wrong in practice.
• A themed product line (same brand, same audience, different designs): one campaign, but each listing gets its own ad group. Related enough to share a budget line, distinct enough to need separated performance data.
• Completely unrelated products: a separate campaign per product, full stop. No shared audience, no reason to share a budget or a keyword list.
• A mixed catalog (some of each): map your sub-groups first, then apply the matching pattern to each one, plus cross-group negative keywords so a search for one sub-group doesn't waste spend triggering an ad from a different one. See cross-cannibalization for what happens when that negation is skipped.

Why this matters more before you launch than after

Restructuring a live campaign isn't just annoying, it's a real cost. You lose the performance history tied to the old structure, and Amazon's algorithm has to relearn your account from a colder start (more on exactly what a rebuild costs you). The sellers who get this decision right on day one aren't smarter, they just had the framework before they needed it.

---

Want the answer for your actual catalog instead of the general framework? The free campaign structure check asks a few questions about your products and tells you exactly which pattern fits, no guessing required.