Amazon Ads vs Google Ads: Two Different Funnels, One Purchase

Amazon Ads is a retail media network — ads inside a store, shown to shoppers already mid-purchase, converting on Amazon's own checkout against first-party purchase data. Google Ads is an open-web platform spanning Search, Shopping, Performance Max, Display, and YouTube, reaching people from the first research query through to the final buy. These are different instruments for different jobs.

The phrase "Amazon Ads vs Google Ads" implies a head-to-head with one winner. That is the wrong frame. WebFX puts it directly: Amazon captures shoppers at the moment of purchase; Google creates the demand that leads to that moment. Most e-commerce brands that sell on Amazon run both — the real decision is the split and how each platform gets measured. New to the Amazon side? Start with what Amazon Ads is.

Amazon Ads runs through three self-service ad types — Sponsored Products, Sponsored Brands, and Sponsored Display — all inside the Amazon store, with conversion on Amazon's checkout rather than your site. The managed tier, Amazon DSP and Streaming TV, extends reach off-Amazon, but attribution still flows back through Amazon's closed-loop first-party data. Google Ads has no equivalent data moat: its attribution is cross-web and model-dependent. Google spans the full funnel — Search, Shopping, Performance Max, Display, YouTube, and Demand Gen — which makes it the demand-generation engine Amazon is not, and is why a head-to-head CVR comparison misleads: Amazon shows ads only to in-market shoppers already in a buying context.

This pillar covers the four dimensions where the platforms diverge — cost (ACOS vs CPA/ROAS, CPC by category), shopper intent and funnel position, attribution (closed-loop first-party vs cross-web modeled), and ad formats — and closes with a budget-split decision framework. All cost benchmarks are vendor-specific: Amazon from Trellis (platform data, March 2026) and Insider Intelligence 2025 via Bridgeway Digital; Google from WordStream/LocaliQ 2026 (13,474 US campaigns, April 2025–March 2026). The two columns in every table use different vendors and are not directly comparable. For the DTC brands in our client mix spanning skincare, supplements, and home goods, the Amazon vs Google allocation question comes up at nearly every quarterly review.

Key Facts at a Glance

Amazon Ads = demand capture (retail media, first-party purchase data, ACOS). Google Ads = demand generation + capture (open web, Search through YouTube, CPA/ROAS). Most e-commerce brands that sell on Amazon run both; the split — not the choice — is the decision.

• CPC gap, sourced separately: Amazon Sponsored Products CPC $0.75–$3.50+ by category (Trellis, Mar 2026); Google all-industry search CPC $5.42 (WordStream/LocaliQ 2026, Google + Microsoft search). Different vendors, different platforms — not directly comparable.
• CVR by context: Amazon Sponsored Products CVR 10–15% (Trellis midpoint 12.5%); Google all-industry search CVR 8.18% (WordStream/LocaliQ 2026). Amazon's higher CVR reflects in-store, purchase-ready audience context — not absolute platform superiority.
• Metric architecture: Amazon efficiency metric = ACOS (ad spend ÷ ad sales; break-even ACOS = your profit margin). Google efficiency metrics = CPA and target ROAS. ACOS and ROAS are reciprocals (25% ACOS = 4× ROAS) but measure completely different conversion surfaces and must not be compared head-to-head.
• Retail media growth: US retail media reaches $69.33 billion in 2026 (+17.9%); Amazon holds ~79.7% of US retail media (eMarketer H1 2026). Google's US search ad share: 48.5% in 2026 — first time below 50% in 20+ years (eMarketer).
• Attribution divergence: As of January 1, 2026, Amazon retired the 14-day view-through window for store ads in favor of a shopping-signal ML last-touch model. Click-through attribution is unchanged. Google uses data-driven attribution (DDA) by default since retiring rule-based models in 2023.
• Hybrid advantage: Advertisers running Amazon and Google together report "up to 42% higher ROI" versus single-channel (Bridgeway Digital, citing Insider Intelligence 2025). For brands that sell on Amazon, both platforms is the default operating mode.

Amazon Ads vs Google Ads: At a Glance

Amazon Ads is a retail media network measuring on-Amazon sales against first-party purchase data, with ACOS as the efficiency headline. Google Ads is an open-web platform measuring off-platform conversions via data-driven attribution and target ROAS or CPA. The table below maps the structural differences across eight key dimensions.

The "which is better" framing fails because the two platforms answer different questions. Amazon is the shelf you win at the moment of purchase — a shopper there is already in a buying context, searching a product category, about to check out. Google is the journey that gets people there: informational queries, comparison searches, YouTube reviews, and high-intent keywords that drive traffic to your site. Amazon and Google serve the same purchase dollar from opposite ends of the funnel.

For ecommerce brands in categories like electronics, furniture, and pet supplies, the choice is rarely "either/or." The real question is how much of a fixed budget goes to Amazon's demand-capture and how much to Google's demand-generation. The cost, CVR, attribution, and format comparisons below inform that split; the Amazon-side execution — bidding, placement modifiers, budget pacing — lives in Amazon Ads bidding and budgets.

Cost Comparison: CPC, CVR, and ROAS by Platform

Amazon Sponsored Products CPC runs $0.75–$3.50+ by category (Trellis, March 2026); Google all-industry search CPC is $5.42 (WordStream/LocaliQ 2026). The gap is an intent gap in disguise, not a bargain: Amazon's lower CPC reflects a narrower, purchase-ready audience, while Google's higher CPC buys reach across the full funnel, including the awareness and research stages Amazon never pays for.

The ROAS comparison above — Amazon 5.2× vs Google 4.1× (Insider Intelligence 2025 via Bridgeway Digital) — is directionally useful but not apples-to-apples: Amazon ROAS counts on-Amazon closed-loop sales, Google ROAS counts modeled off-site conversions. Stacking the two to declare a winner is a category error. For full ACOS↔ROAS↔CPA translation, see Amazon Ads ACOS, ROAS, and metrics.

A cost-per-conversion estimate from the vendor figures: Amazon Sponsored Products at $0.89 CPC and 9.55% CVR (Insider Intelligence 2025 via Bridgeway) is about $9.32 per Amazon conversion; Google search at $5.42 CPC and 8.18% CVR (WordStream/LocaliQ 2026) is about $66.26 per Google conversion. Amazon's implied cost looks dramatically lower, but the two measure different things — an on-Amazon retail sale with first-party data vs an off-platform site conversion across a full-funnel audience. The comparison shows the intent gap, not a reason to move budget without weighing what each job is worth.

Amazon CPC by Product Category (Trellis, March 2026)

Amazon Sponsored Products CPC ranges from $0.70 for Toys & Games to $1.50 for Beauty & Personal Care and Grocery — well below Google's all-industry search average of $5.42. The chart shows category midpoints against the Google blended reference line. These are different vendors measuring different platforms; the gap reflects intent, not a simple price advantage.

Several patterns emerge from the Trellis data. Categories with high product discovery — Beauty & Personal Care, Grocery — show higher CPCs because shelf competition is intense and the purchase occasion is frequent. Categories where price and spec comparison drives search — Electronics, Tools — also trend above the platform average, though Trellis does not publish electronics as a separate line because its variation is too wide to state as a range. For brands running kitchen appliances PPC or sports equipment PPC on Amazon, category CPC is a starting point for bid planning, not a guarantee.

The Google $5.42 all-industry figure (WordStream/LocaliQ 2026) has no per-Amazon-category equivalent — the two providers cover different platforms and taxonomies, and a category-level Amazon vs Google CPC comparison sourced to one vendor does not exist publicly. The chart shows Amazon category midpoints alongside Google's blended reference as a labeled single-bar reference, not a matched comparison. For beauty products PPC or baby products PPC campaigns, both budgets inform the overall CAC stack — not substitutes, but complementary capture and generation spend.

Conversion Rate: Amazon Ad Formats vs Google Ads

Amazon Sponsored Products converts at 10–15% (Trellis midpoint 12.5%); Google all-industry search CVR is 8.18% (WordStream/LocaliQ 2026). Amazon's higher CVR is structural: it shows ads only to shoppers already in a purchase-ready retail context, while Google covers the full funnel, including awareness and research stages that convert lower by design.

The Bridgeway Digital head-to-head (Insider Intelligence 2025) reports Amazon CVR at 9.55% vs Google at 3.75%. That lower Google figure (vs WordStream's 8.18%) reflects a blended campaign-type mix or different date range — the methodology is not fully disclosed. For the Google column above, WordStream/LocaliQ 2026 (13,474 US campaigns, April 2025–March 2026, medians) is the preferred figure: the larger, more recent, clearly-stated dataset covering search-only campaigns. When citing the Bridgeway head-to-head, use its stated 3.75% with full attribution.

The practical insight: Amazon's high CVR reflects purchase-ready in-store intent, while Google's CVR spans a far wider intent range — high-intent search queries convert near or above Amazon's rate, broad awareness queries well below it. Comparing one blended CVR from each platform without accounting for intent mix is how brands reach the wrong conclusion about which "performs better." For brands in jewelry PPC or shoes PPC, Google Shopping CVR — product-level comparison intent — runs materially higher than Google's blended search rate.

Intent and Funnel Position: Where Each Platform Operates

Amazon Ads is dominant at purchase capture — shoppers who arrive on Amazon are already mid-buy, and Sponsored Products intercepts them at the category or branded-keyword level. Google Ads covers awareness, consideration, and purchase capture across the open web via search queries, shopping signals, and behavioral audiences. The two do not compete at the same funnel stage.

The customer-relationship row is the strategic fact most brands underweight in an "Amazon only" approach. On Amazon, the shopper is Amazon's customer — the brand cannot access the email, build a retargeting list from purchase data (without Amazon DSP custom audiences), or own the repeat-purchase journey. On Google, every click that lands on your site is a relationship the brand owns: email capture, first-party pixel data, retargeting, and loyalty. For categories where lifetime value matters — skincare PPC, supplements and nutrition PPC, home goods PPC — the long-term argument for Google is the relationship asset, not just the conversion.

Targeting granularity follows the same split. Amazon's self-service Sponsored ads target at the marketplace and locale level — no city or radius parameter in the Sponsored Products or Brands console, and granular geo exists only inside Amazon DSP. Google offers city, radius, and DMA targeting on every campaign type, so brands that need local-intent capture cannot replicate Google Local with Amazon's self-service tier. Amazon's targeting options are purpose-built for retail media: keywords, ASINs, product categories, and purchase- and browse-based audience segments.

Ad Format Comparison: Amazon Ads vs Google Ads

Amazon Ads has five main format tiers: Sponsored Products, Sponsored Brands, Sponsored Display (self-service CPC/vCPM), Amazon DSP (managed programmatic CPM), and Streaming TV (CTV CPM). Google Ads has six format families: Search, Shopping, Performance Max, Display (GDN), YouTube, and Demand Gen. The overlap is full-funnel automation: Amazon's Performance+ and Google's Performance Max both use ML, but on entirely different data and conversion surfaces.

A naming note: Amazon's current brand is Amazon Ads (not "Amazon Advertising" or the legacy "AMS"). Never write "Performance Max" for Amazon — its equivalent automated format is Performance+, which operates within Amazon DSP. Conflating the two causes platform-mechanic errors in setup and reporting.

The format split also clarifies the audience-ownership point from the funnel table. Every Amazon format — SP, SB, SD, DSP, STV — converts inside Amazon's closed ecosystem, so the brand's customer data stays with Amazon. Every Google format drives traffic to a URL the brand controls, where customer-data collection is the brand's asset. For brands in electronics PPC or home goods PPC running both an Amazon presence and a DTC site, format strategy maps directly to channel-ownership strategy.

Retail Media Market Context: The Budget Shift

US retail media ad spending reaches $69.33 billion in 2026, up 17.9% from $58.79 billion in 2025, with Amazon holding ~79.7% of the US retail media market. Over the same period, Google's US search ad share falls below 50% for the first time in over 20 years — a structural shift, not a substitution, with both pools growing and Amazon capturing the incremental dollar.

The market context also explains why the "Amazon is eating Google's lunch" narrative is half right. Amazon's US search ad revenues are growing 17.6% vs Google's 7.6% (eMarketer 2026) — Amazon is taking search-intent budget because shoppers increasingly start product searches on Amazon. But Google still reaches shoppers earlier, dominates non-product informational search, and runs the only full-funnel platform that converts traffic to your own site. The answer for most ecommerce brands is to fund each for the job it does and measure it on that job's own metric, not to choose by growth rate.

For brands managing both channels, the frame is simple: retail media is growing faster, and Google remains dominant in open-web search and brand building. For brands in home goods PPC, furniture PPC, or pet supplies PPC, build Amazon spend as a baseline that captures existing demand and use Google to expand the addressable market and generate demand that flows through Amazon or your DTC site.

Attribution: Closed-Loop First-Party vs Cross-Web Modeled

Amazon attribution is closed-loop and first-party — it sees the actual purchase because it owns the checkout. As of January 1, 2026, Amazon retired the legacy 14-day view-through window for store ads and DSP vCPM billing for a shopping-signal ML last-touch model. Google uses data-driven attribution (DDA) by default, ML-weighted across all touchpoints, since retiring rule-based models in 2023.

The January 2026 Amazon attribution change matters for advertisers comparing ROAS pre- and post-change. Under the old model, any ad view within 14 days of a purchase could claim credit; under the new ML model, only views at "brand discovery moments" — exploratory browse and category search — receive credit. View-through ROAS for Sponsored Display and DSP appears 15–30% lower in 2026 reporting (PPC.land), and Code3-managed DSP advertisers see "double-digit declines in attributed DSP view-through revenue." These are measurement changes — actual sales are not declining; the legacy "all views" 14-day metric stays accessible in Unified Reporting (labeled "Purchases (all views)"). Full detail on Amazon's attribution mechanics is in Amazon Ads ACOS, ROAS, and metrics.

Google's data-driven attribution is structurally different: it assigns credit probabilistically across every touchpoint, using machine learning to weight each interaction by its observed contribution to conversion. A YouTube awareness view, a Display remarketing impression, and a final Search click all receive fractional credit — not just the last click. DDA is also cross-device, stitching the journey across phones, tablets, and desktops via signed-in accounts, and imports off-platform conversions via Offline Conversion Tracking. The picture is broader but modeled, dependent on tracking quality. Amazon's attribution is narrower (on-Amazon, Sponsored ads) but more certain. Neither is "more accurate" — they are accurate about different things.

When to Use Each Platform and How to Split Budget

For most ecommerce brands that sell on Amazon: run both platforms, set a separate efficiency target for each, and measure them independently. Amazon's target is break-even ACOS (your profit margin). Google's target is a CPA or new-customer acquisition cost your LTV model supports. Bridgeway Digital cites a 60/40 Amazon-to-Google split as one illustrative starting point for a sale-focused brand; invert for DTC-first brands.

MB Adv Agency applies a clear rule when setting initial splits: fund Amazon to its break-even ACOS first, then direct the rest to Google for the demand-generation and brand-defense work Amazon cannot do. The review is always quarterly and uses each platform's own metric — never a side-by-side ROAS comparison of two different measurement surfaces. For brands running jewelry PPC or shoes PPC, the split logic holds even when the absolute allocation differs by category CPC and margin.

The hybrid ROI premium from Bridgeway (Insider Intelligence 2025) — "up to 42% higher ROI" vs single-channel — is attributed to that vendor, not mbadv client data. Directionally it reflects demand-gen (Google) feeding demand-capture (Amazon): Google creates new-to-brand awareness that converts on Amazon, which captures it with lower friction and higher CVR. The two platforms are additive, not competitive. For budget-allocation mechanics on the Amazon side, see Amazon Ads bidding and budgets.

Three Misconceptions About Amazon Ads vs Google Ads

The "Amazon Ads vs Google Ads" framing invites three recurring errors that cause brands to mis-allocate budget and compare incompatible metrics. Each is refutable from the data above.

Misconception 1: "It's Amazon or Google — pick the cheaper, higher-converting one."

Amazon and Google are not substitutes competing for the same click. Amazon captures shoppers at the point of purchase using first-party data; Google generates and captures demand across the open web, with conversion on your own site. Picking "the cheaper one" on CPC ignores that Amazon's lower CPC reflects a narrower, purchase-ready audience (WebFX; Bridgeway Digital 2026), while Google's higher CPC buys reach across stages Amazon never pays for. MB Adv Agency corrects this framing with new ecommerce clients: optimizing against the wrong comparison drives under-investment in whichever platform does not look cheaper on its own metric.

Misconception 2: "Amazon's lower CPC and higher CVR prove it's the better platform — shift budget there."

Lower CPC and higher CVR are structural features of a closing-stage retail-media channel, not evidence of platform superiority. Amazon shows ads only to in-market shoppers already on a purchase mission, so its clicks are cheaper and convert more often. Scale Insights and Bridgeway Digital both note this context. Google pays to work the research, comparison, and awareness stages Amazon never touches — the work that creates the demand Amazon later captures. The correct question is whether Amazon is hitting its ACOS target and Google its CPA or new-customer target. If both are yes, fund both more and put the next dollar where it generates the most incremental value.

Misconception 3: "Amazon ROAS and Google ROAS are the same number — compare them directly."

ACOS and ROAS share a formula; Amazon ROAS and Google ROAS do not share a measurement universe. Amazon ROAS counts on-Amazon sales inside a closed first-party loop — it sees the actual purchase because Amazon owns the checkout. Google ROAS counts conversions your tags observe on your own site, modeled cross-device via data-driven attribution. The Bridgeway/Insider Intelligence 2025 head-to-head (Amazon 5.2× vs Google 4.1× ROAS) is a directionally useful single-vendor read, but stacking the two figures to declare a winner is a category error (ecomBrainly; Google Ads Help). Map ACOS↔ROAS↔CPA to understand each platform individually — then stop.

ACOS, ROAS, and CPA: Translating Metrics Across Platforms

ACOS and ROAS are reciprocals: 25% ACOS = 4× ROAS, and break-even ACOS equals your pre-ad gross profit margin. Google's equivalent is CPA, set against a target ROAS in smart bidding. The translation is mathematical, but the measurement surfaces differ — and treating them as equivalent is the source of most cross-platform budget errors.

MB Adv Agency finds the brands that struggle most with cross-channel allocation are those comparing Amazon ROAS and Google ROAS side by side without accounting for what each number measures. Amazon ROAS counts on-Amazon sales in a closed loop; Google ROAS counts modeled cross-device conversions on your own site. Setting one ROAS target and expecting both platforms to hit it is a structural mistake.

The translation workflow for an omnichannel brand:

1. Identify your pre-ad gross margin per product — it sets Amazon's break-even ACOS.
2. Set a target ACOS at or below break-even for Sponsored Products; for new-category entry, accept a higher ACOS to buy share and track TACOS (total ad spend ÷ total Amazon revenue including organic).
3. Set a separate Google CPA target from your LTV model — new-customer CPA, not blended, since returning customers inflate the figure.
4. Run each platform to its own target and rebalance quarterly, never by the gap between the two ROAS numbers.

For brands in baby products PPC or kitchen appliances PPC, the ACOS targets on Amazon and CPA targets on Google are category-specific — set by category CPC, CVR, and product margin, not a platform-agnostic benchmark. The full metric framework for the Amazon side is in Amazon Ads ACOS, ROAS, and metrics.

Frequently Asked Questions

The questions below cover the core decision points for ecommerce brands evaluating Amazon Ads, Google Ads, or both — from cost translation to attribution mechanics to budget split logic.

Methodology

Cost benchmarks (CPC, CVR, CTR, ROAS, ACOS) are sourced from platform-specific vendors: Amazon figures from Trellis platform data (March 2026) and Insider Intelligence 2025 via Bridgeway Digital; Google figures from WordStream/LocaliQ 2026 (13,474 US campaigns, Google + Microsoft search, April 2025–March 2026, medians). Retail media market figures from eMarketer H1 2026 forecast and US Search Advertising Forecast 2026. Attribution model details from Amazon Ads official documentation (January 2026 model change), PPC.land, Code3, and Google Ads Help. No Amazon and Google figures share a vendor — the two platforms are not tracked by equivalent providers — and Amazon figures are never cross-applied to Google columns or vice versa. No mbadv client metrics are cited; all agency POV is qualitative. Reviewed by MB Adv Agency, June 2026.