Been testing ad structures for 7 years, and this is the one setup I keep coming back to.

I’ve run Amazon PPC for 7+ years across 20+ categories, and for the longest time I struggled with two huge pain points: broad match ads wasting thousands on irrelevant traffic, and exact match campaigns that never seemed to move the needle on organic rank.

Earlier this year I tested a tiered broad-to-exact funnel framework for 12 client ASINs, and the results were consistent across every single one: average ACOS dropped 22% in 30 days, and core keyword organic rank improved 3x faster than with our old ad structure. I’m breaking down exactly how it works today, no fluff.

First, let’s clear up the core roles of broad and exact match, because most sellers get this backwards. A lot of new sellers treat broad match as a set it and forget it way to get traffic, and exact match as a way to lock in low ACOS once you have top rankings. That’s the opposite of how I use them to build scalable, profitable campaigns.

Broad match is your sifting tool. Amazon’s broad match algorithm pulls in every search term related to your target keyword, including synonyms, variant phrases, and rearranged word orders that you’d never think to target manually. That makes it perfect for uncovering hidden high-converting long-tail terms, especially in categories with tons of attribute combinations like apparel, home goods, or toys.

Exact match is your rank-pushing tool. It only triggers when a shopper searches for your exact target keyword (ignoring only minor differences like plurals or prepositions), so every single click and conversion gets applied directly to that specific keyword’s weight in Amazon’s A9 algorithm. That concentrated weight makes it the fastest way to push a core keyword’s organic search position up.

Let’s use a real example from a client’s holiday line: a girls’ Christmas dress. If you’ve ever sold apparel, you know how many search variations there are for a single product. You’ve got size attributes (size 8, toddler 4T), style attributes (long sleeve, lace, tulle), occasion attributes (Christmas party, holiday pageant), even character attributes (Elsa, Disney princess). There’s no way you can list every single possible combination manually when you launch.

Broad match solves that. If you target girls Christmas dress on broad match, Amazon will pull in all those random long-tail searches for you. The problem most sellers run into is that all that traffic gets mixed together, leading to overlapping spend between campaigns and wasted budget on terms that don’t convert.

That’s where the tiered funnel and layered negative targeting comes in. First, you’ll want to sort your core keywords into three tiers based on attribute depth:

Tier 1 (core root term): girls Christmas dress (we avoid the even broader girls dress to keep initial traffic relevant to our holiday product)

Tier 2 (core + 1 attribute): girls Christmas dress size 8

Tier 3 (core + 2+ attributes): girls Christmas dress size 8 long sleeve

For each of these three tiers, you’ll create two separate campaigns: one broad match campaign for sifting new terms, and one exact match campaign for pushing rank. That gives you 6 total core campaigns for your core keyword root.

To stop internal competition and make sure each broad match campaign only pulls in new, unique terms for its tier, we use a bottom-up negative targeting rule:

  • Add all Tier 3 keywords as phrase negatives in the Tier 2 broad match campaign

  • Add all Tier 2 and Tier 3 keywords as phrase negatives in the Tier 1 broad match campaign

  • Add exact match negatives for each keyword in its own broad campaign to avoid overlap with the exact match rank-pushing campaign

Let me break down why this works. If you don’t add Tier 3 terms as phrase negatives to the Tier 2 broad campaign, every search for girls Christmas dress size 8 long sleeve red would get pulled into both the Tier 2 and Tier 3 broad campaigns, doubling up on spend for the same search. With the negation rule in place, that search only goes to the Tier 3 broad campaign, so the Tier 2 broad campaign can focus on uncovering other Tier 2 attribute combinations like girls Christmas dress size 8 red or girls Christmas dress size 8 lace that you haven’t already identified.

Same for the Tier 1 broad campaign: once you negate all Tier 2 and 3 terms, it only pulls in completely new root variations like girls Christmas party dress or toddler girls Christmas dress that you haven’t already categorized into lower tiers. This turns your broad match campaigns into a perfectly calibrated funnel that sifts out new high-potential terms without wasting budget on terms you’re already targeting.

I’ve tested this with apparel, home decor, and even pet supplies, and it consistently cuts wasted broad match spend by 30-40% in the first two weeks of implementation. You can pull your search term reports weekly via Seller Central, or use tools like Helium 10 or Jungle Scout to automate sorting high-performing terms into new tiers as you find them.

Now let’s talk about the exact match side of the funnel, because this is where you turn those sifted high-performing terms into organic rank that drives free sales long term.

The reason exact match is the only efficient tool for pushing rank comes down to how Amazon’s A9 algorithm weights keyword relevance. Every time a shopper searches for a specific keyword, clicks your ad, and converts, that conversion adds weight to your listing’s relevance for that exact keyword. With broad match, those conversions get spread across dozens of different search terms, so you never build enough concentrated weight for any single term to move the needle on organic rank.

With exact match, 100% of your conversions get applied directly to your target keyword. For example, if you’re targeting girls Christmas dress size 8 on exact match, every conversion from that campaign counts directly toward that keyword’s relevance score. That’s why exact match campaigns can push a keyword from page 10 to page 1 in as little as 4 weeks if you’re hitting your conversion targets.

The most efficient spot to run your exact match rank-pushing campaigns is the Top of Search (TOS) position on the first page of results, since that’s where click-through rate (CTR) is highest. The catch here is that TOS positions have higher Cost-Per-Click (CPC) costs, so you need to make sure your listing is optimized to convert before you pour budget into this step. If your review count is lower than top competitors, run a limited time Amazon Coupon to boost conversion rate to match or beat competing listings in that position.

Let’s walk through the full rank-pushing loop so it’s clear:

  1. First, you lock in your target high-performing keyword (sifted from your broad match funnel) in an exact match campaign.

  2. Next, you bid enough to win TOS placement to get consistent high exposure and clicks.

  3. Then, you optimize your listing and offer to keep conversion rate at or above the category average for that keyword.

  4. As you generate consistent conversions from the exact match campaign, Amazon’s algorithm picks up that your listing is highly relevant and popular for that keyword, and starts boosting its organic search position.

  5. Once your organic rank hits the top half of page 1, you can gradually lower your exact match bids, since you’ll start getting most of your traffic and sales from the free organic position. That’s how you turn ad spend into long-term, low-ACOS revenue.

For that Christmas dress client I mentioned earlier, we sifted the term girls Christmas dress size 8 long sleeve from the Tier 3 broad campaign in week 2, moved it to an exact match campaign targeting TOS, and pushed its organic rank from position 47 to position 3 on page 1 in 32 days. We then lowered exact match bids by 60%, and the keyword still drove 18 organic sales per day for the rest of the holiday season with an ACOS of only 8%.

To wrap this up into a simple, repeatable workflow you can apply to any of your ASINs:

  • Build your initial tiered keyword list for your core root term, starting with 1 Tier 1 term, 2-3 Tier 2 terms, and 3-5 Tier 3 terms based on initial keyword research.

  • Launch the 6 core campaigns (broad + exact for each tier) with the bottom-up negative targeting rules in place to avoid overlap.

  • Pull search term reports weekly to identify new high-converting terms (aim for terms with CTR > 0.4% and CVR > 1.5x your category average) and sort them into the appropriate tier, or add them as new tiers as you go.

  • Move high-performing terms that are consistent converters into their own exact match campaigns, target TOS placement, and optimize your listing to maintain strong conversion rates.

  • Once organic rank for a term hits page 1, gradually lower exact match bids to reduce ad spend while keeping the organic sales coming in.

A quick note on CPC costs: these numbers will vary widely by category and site, so always adjust your bids based on your specific marketplace’s average CPC. For example, CPCs in the US apparel category average $1.20-$1.80, while CPCs in the EU home goods category may run 30-40% lower. Always reference your category’s average benchmark before setting bids.

Warning: Avoid adding too many terms to your exact match campaigns too quickly. Start with 2-3 top performing terms first, make sure you have the budget to support consistent TOS placement, and add more terms as you start seeing organic rank improvements. Spreading your budget too thin across 10+ exact match terms will slow down rank growth for all of them.

I’ve used this framework across 7 different Amazon marketplaces, and it works for everything from low-cost consumables to high-ticket electronics. The only tricky part I’ve run into is handling multi-root keywords with overlapping attributes, which takes a bit of extra tweaking to the negation rules.

Curious if anyone else has seen similar results—or if this setup flops for your category. Let me know in the comments!

Have you tried a tiered broad-to-exact funnel for your PPC campaigns? Did you run into issues with overlapping traffic or slow rank growth? Drop your experience or questions below, I’ll answer every comment.