PPC campaign architecture

Amazon's ad platform spans three ad products — Sponsored Products (SP), Sponsored Brands (SB/SBV), and Sponsored Display (SDI/SDV) — and every campaign you run should have a declared purpose before it launches. Purpose determines bid strategy, match type, budget allocation, and how you interpret the data it produces.

There are four purposes: Ranking, Performance, Research, and Shielding. The campaign architecture below maps each ad type to those purposes and explains how they layer across a product's lifecycle.

Sponsored Products (SP)

SP is the foundation. It generates the most clicks, the most purchase data, and — when structured correctly — the strongest ranking signal Amazon receives.

Auto campaigns — Research

SP Auto campaigns let Amazon decide which keywords and products to show your ads against. They run across four targeting sub-types: Close Match, Loose Match, Substitutes, and Complements. Each sub-type is best run as a separate campaign so you can set independent bids and negate targets without affecting the others.

Auto campaigns are no longer just research tools. Amazon's relevance algorithm has improved significantly — in many accounts, an auto Close Match campaign is the top-performing campaign by both sales volume and ACoS. Start there. Add Substitutes once you've mined Close Match for a few weeks.

Negative targeting is mandatory. Exact-negate all keywords and ASINs you're already targeting in manual campaigns. Without negatives, auto campaigns cannibalize spend from your precision campaigns and muddy your bid signals.

Audience bid adjusters are a recent addition to Sponsored Products campaigns. When enabled, Amazon increases your bid for shoppers with high purchase likelihood based on recent shopping activity. Across multiple accounts, audience-targeted impressions consistently produce ACoS 10-20 percentage points lower than account averages. Consider adding this adjuster to every SP campaign — it costs nothing when it doesn't fire, and meaningfully reduces ACoS when it does.

Manual keyword campaigns — Performance and Ranking

Manual keyword campaigns target specific queries at specific match types. The distinction between Performance and Ranking purpose matters here:

• Performance campaigns target keyword roots (e.g., "exfoliating toner pads") in Exact and Broad match across multiple root groups. Goal: profitable sales at or below target ACoS. Bid strategy: Dynamic Down Only. Run multiple root-group campaigns rather than one large campaign — it makes optimization and reporting much cleaner.
• Ranking campaigns target one keyword (or a tight semantic cluster) in Exact match with placement modifiers tuned to push 80% or more of clicks to Top-of-Search. The starting bid formula is: target ACoS × (product price × conversion rate). These campaigns are expensive by design — you're paying for rank velocity, not immediate profitability.

Broad Modified (Broad/M) sits between broad and phrase. It requires all modified terms to appear in the query in any order. Useful for consumables with high-volume root keywords where phrase is too restrictive.

Manual product targeting campaigns — Performance

Instead of targeting queries, you target specific competitor ASINs or categories. In 2026, product targeting campaigns are frequently the highest-selling and most profitable campaigns in an account — they show in both search results and on product detail pages.

Three standard product targeting campaign types:

• Lower Rated ASINs — competitors with a lower star rating than your product. Shoppers comparing products are a captive audience.
• Higher Priced ASINs — competitors priced above you. Your lower price is the conversion advantage.
• Missing Feature ASINs — competitors lacking something your product has. Target the feature gap.

Self-Targeted Product Placement (STPP) is a special fourth type: you target your own ASIN. It functions as immediate retargeting — capturing shoppers already on your listing who haven't converted yet. This is the only SP campaign type that uses Dynamic Up and Down bidding (all others use Down Only).

Category targeting with refinements — Performance

Instead of targeting individual ASINs, category targeting lets you target an entire product subcategory. The power move is adding refinements: filter by products with star ratings under a certain threshold, or products priced above a specific amount. This ensures your ad only appears on listings where your product has a visible advantage — better rating, lower price, or both.

Category targeting often delivers the lowest ACoS of any campaign type in an account. It's lean by design: you define the constraints, and Amazon finds the right placements within them.

Sponsored Brands (SB/SBV)

SB ads occupy a different placement than SP — they appear at the very top of search results, above even the top-of-search SP placement. Clicks and purchases from SB and SBV campaigns do contribute to organic keyword ranking.

Sponsored Brands — Product Collection

Product Collection ads show a headline, your brand logo, and up to three products. They link either to a landing page (a subset of your catalog on Amazon) or to your Brand Store. Use keyword targeting with Exact match for ranking-purpose campaigns, Broad for wider performance reach.

Sponsored Brands Video (SBV)

SBV is the standout format of 2025–2026. These are inline video ads that play in search results. They can be keyword-targeted (ranking/performance) or product-targeted — and the combination of SBV format with product targeting is one of the highest-performing campaign configurations available right now.

Creative requirements: 15–30 second video, shows the product in use, first 3 seconds must communicate the problem being solved or the core benefit. CTR is driven by the first three seconds. Everything else follows from that.

The SBV + Product Targeting combination is the standout campaign configuration of 2026. Combining the SBV ad format with product targeting as the targeting method rides two trends simultaneously — video creative dominance and product targeting's improved performance. In accounts that run this combo, it's frequently the top campaign by sales volume with strong ACoS. If you have video creative, test this combination before scaling other campaign types.

SBV campaigns use CPC bidding with placement optimization toward Top-of-Search. Download the SB placement report periodically and adjust placement modifiers to keep the bulk of spend on top-of-search placements.

Sponsored Display (SDI/SDV)

SD ads appear outside the core search experience — on product detail pages, in browse widgets, and off Amazon via the Amazon DSP network. They are awareness, retargeting, and competitor-conquest tools.

Three optimization goals:

• VCPM (viewable cost per mille) — awareness, run toward competitor and category pages
• PV (page visits / traffic) — driving clicks back to your listing
• CNV (conversions) — retargeting shoppers who viewed or purchased

Product targeting (SDI/SDV)

Target specific competitor ASINs or category pages. Combined with video creative, this is the fastest way to occupy every visible ad slot on a high-traffic competitor listing.

Remarketing

• Views remarketing — retarget shoppers who viewed your product but didn't buy. Standard lookback: 14–30 days. For high-priced products, extend to 60 days — consideration windows are longer.
• Purchase remarketing — retarget past buyers. High ROI for consumables where the repurchase cycle is the business model.

Standard products can skip SD without major consequence. Consumables and high-priced items should include at least views remarketing.

Newer targeting formats

Keyword theme targeting

Amazon now offers keyword theme targeting (formerly called "shopper intent") as a targeting option in Sponsored Products. Instead of targeting individual keywords, you target clusters of related search intent. Amazon manages the keyword expansion within the theme. This format works well for broad awareness and for accounts moving away from rigid exact-match-only structures.

Bonus strategies

Scavenger campaigns

Set a very low bid — roughly one-tenth of Amazon's suggested bid — across a large number of keywords or product targets. These campaigns wait for competing campaigns to exhaust their daily budgets, then capture the now-cheap impressions. The result is very lean ACoS on sales that competitors left on the table. Scavenger campaigns won't drive volume on their own, but they're a low-risk way to pick up incremental sales at minimal cost.

Putting it all together

Campaign architecture follows the product lifecycle:

Launch (under 25 reviews): Lead with Research (auto Close Match and Substitutes) and product-targeting Performance campaigns against weaker competitors. Hold back aggressive ranking campaigns — social proof matters for conversion and low review counts hurt CVR, which undermines the ranking signal.

Build (25–75 reviews): Add Ranking campaigns on lower-competition keyword roots. Broad match Performance campaigns across 2–3 keyword root groups. Brand keyword shielding once you have established branded search volume.

Scale (75+ reviews): Full architecture: multiple Ranking campaigns on priority keywords, all root groups in Exact and Broad, full product targeting suite (lower rated, higher priced, STPP), all four auto sub-types, SBV for both ranking and performance, SD for category/competitor targeting and views remarketing.

Budget split benchmarks:

• SP typically absorbs 60–85% of total daily budget
• SB/SBV: 10–30%
• SD: 0–25% (skew toward SD for consumables and high-priced products)

The platform is shifting from narrow to broad. Over the past 12-18 months, Amazon's ad system has moved from narrow exact-match targeting being dominant to broader targeting — category targeting, auto campaigns, broad match, keyword themes — performing best for many accounts. This mirrors a shift that already happened on Meta (2022-2023) and Google (shortly after). The implication: don't over-index on single-keyword exact-match campaigns at the expense of broader discovery and product targeting.

Naming convention: Every campaign name should encode its purpose so you can filter and report without opening individual campaigns. The standard pattern is {Initials} | {Product} | {ASIN} | {Market} | {AdType} | {Targeting}. RankASIN's campaign blueprint uses this pattern and stores company_initials, marketplace, and product_nickname on the seller account to auto-generate names.