Retail Media ROAS Benchmarks by Industry & Platform, 2026

Last updated: October 2026 · Reviewed by Kunal Damgude, Growth and Product Marketing Manager at Osmos

A good retail media ROAS (return on ad spend) in 2026 is one that beats what a brand earns from search and social on the same attribution basis. The closest thing to a market average is Skai's Q1 2026 figure, tracked in Osmos's ROAS benchmarks: retail media returned $6.00 in attributed sales for every dollar of ad spend, holding at that level through a fifth consecutive quarter of rising spend. Category returns spread far wider than the average,.

Osmos is a white-label retail media operating system for retailers and marketplaces. Its modular apps for onsite, offsite and in-store ads sit over your existing stack, so you can break your monetisation ceiling faster without rebuilding it.

Series context. This is the hub for Osmos's ROAS and performance benchmarks series. The spokes carry the format, channel and economics depth: Retail Media Benchmarks 2026: What Success Looks Like, Retail Media Ads: Sponsored Formats 2026, Native Advertising Retail Media Benchmarks 2026, Onsite vs Offsite Retail Media Ad Formats Framework 2026, How Retail Media Offsets Customer Acquisition Costs and How Retailers Use Retail Media to Boost Margins.

What is a good ROAS in retail media in 2026?

A good retail media ROAS is the return that clears a brand's next-best use of the same budget, measured with the same attribution window, and in 2026 the market reference point for that comparison is Skai's $6.00 blended figure for Q1 2026. Skai's chart of the series, as tracked in Osmos's ROAS benchmarks, shows retail media ROAS at $6.05 in Q1 2025, $5.89 in Q2 2025, $6.23 in Q3 2025 and $6.45 in Q4 2025 before returning to $6.00 in Q1 2026 (Skai, Q1 2026 Digital Advertising Trends Report, April 2026). Skai's Q2 2026 release, published in July, reports spend and cost per click but carries no ROAS figure, so $6.00 is the latest blended number on record.

Every figure in the table is advertiser-side, the return a brand earned per dollar on a network it does not own, and the figure a retailer is most likely to hear quoted back at a renewal meeting.

Osmos's 2026 ROAS benchmarks, ordered from the market-wide click price down to single-network returns, compiled from Skai, Pacvue (as summarised by beBOLD Digital) and Teikametrics:

A retailer's Head of Retail Media reads the same numbers the other way round: as the return its own network has to clear before a brand renews, and as the ceiling on what its inventory can be priced at.

"Best ROAS" is therefore measured against the brand's other options for the same budget. A beauty brand earning $3.13 on Amazon and $3.50 on a grocery network, measured the same way, has found the better return, even though both sit well below the $6.00 blended average.

What is ROAS, and how is it calculated?

ROAS is the revenue attributed to an advertising campaign divided by what the campaign cost, written as a dollar figure or a multiple: ROAS = attributed revenue / ad spend. A campaign that cost $1,000 and was credited with $4,200 in sales has a ROAS of 4.2x, or $4.20 per dollar. Osmos reports ROAS this way inside the campaign console, attributed back to the SKU, so a retailer's brands read the same formula the benchmarks use.

A retailer acts on the inputs to that formula: ROAS = conversion rate x average order value / CPC. A click that costs $1.00, converts 10% of the time and produces a $50 basket returns $5.00, or 5.0x. Osmos's ROAS benchmarks use this identity to show which input a network controls, and the biggest of those inputs is the price of a click.

Retail media ROAS is return on ad spend for advertising a retailer offers on its own site, app and stores, and off its properties through audience extension, attributed against the retailer's own transaction data. Compared with open-web ROAS, the ad sits close to the point of purchase, targeting runs on deterministic first-party purchase data, and attribution can include in-store sales where closed-loop attribution is in place. Osmos builds this attribution into each campaign so a retailer can report it without a separate study.

Incremental ROAS (iROAS) is the return counted only on sales that would not have happened without the ad, measured against a holdout group that did not see it: iROAS = incremental revenue / ad spend. Skai's worked example shows the gap: a campaign reporting $4.50 ROAS, where 30% of conversions would have happened anyway, has an iROAS of $3.15 (Skai, "What is iROAS", updated November 2025). Osmos Incrementality, in early access, builds that holdout into the campaign itself so a retailer can report iROAS next to ROAS.

Advertising cost of sales (ACoS) is the inverse of ROAS: ad spend divided by attributed sales, expressed as a percentage. A 25% ACoS is a 4.0x ROAS, and marketplace tooling on Amazon and Walmart reports both, so a brand's team may arrive at a renewal quoting either. Osmos's benchmarks convert ACoS figures to ROAS so that both sides of a renewal compare like with like.

GMV monetization rate is a retailer's retail media revenue as a share of its gross merchandise value (GMV), the network-level counterpart to the campaign-level ROAS a brand sees. Osmos treats it as the headline number for a retailer's own business, because a network can lift advertiser ROAS and still leave revenue on the table if it fills too little of its inventory.

What are retail media ROAS benchmarks by platform, 2024 to 2026?

Platform ROAS in 2026 is best read as a direction of travel: Osmos's ROAS benchmarks show Amazon holding efficiency while it grows and Walmart Connect recovering from a 2025 in which its ROAS fell about 18%, while both posted strong advertising growth in Q2 2026. The market behind them is still expanding. Retail media is a $200.4 billion global market in 2026, rising to $223.4 billion in 2027, and outside Amazon WARC forecasts growth of 9.8%, the lowest rate it has recorded (WARC, via The Desk, Aug 2026). In the United States, advertisers spent $60.32 billion on retail media in 2025 and will spend $71.09 billion in 2026, growth of 17.8%, with Amazon and Walmart capturing 89% of the incremental spend (eMarketer, Jan 2026). Forrester projects global retail media rising from $184 billion in 2025 to $312 billion by 2030, an 11% compound annual growth rate (Oct 2025).

Amazon Ads holds 79.7% of US retail media spend and Walmart Connect 8.0% (eMarketer, Jan 2026). eMarketer's mid-year forecast describes how that share is settling: "Retail media is concentrating as a scaled second tier takes shape and the long tail slips further behind" (eMarketer, May 2026). Each platform's ROAS reflects its own shopper intent, category mix and attribution window, and how fast advertiser demand is bidding up its inventory.

Platform ROAS performance overview, 2025 to 2026

Osmos's platform ROAS and growth data, 2025 to 2026, ordered by share of US retail media spend, from Teikametrics, Tinuiti, Marketing Brew, Skai and the companies' Q2 2026 releases (Amazon, Walmart):

Attribution windows, view-through rules and category mixes differ by platform, so the rows compare on the direction of change. Walmart's 2025 ROAS decline came in the year its spend grew fastest and its click costs rose, according to Teikametrics' data as tracked by Osmos, and by Q1 2026 its Sponsored Products clicks were growing nearly as fast as spend. On both networks, advertiser demand grew faster than efficient inventory in 2025, and a retailer's own network meets the same pressure once its auction fills.

The cost-per-click line, 2024 to 2026, and what it does to ROAS arithmetic

Retail ad cost per click has barely moved from a dollar in seven years: Skai's Q2 2026 release, tracked in Osmos's ROAS benchmarks, puts retail media CPC near a dollar for the seventh straight year even as spend rose 26% year over year (Skai, Jul 2026). Skai's Q1 2026 report showed the same pattern a quarter earlier, with spend up 27%, clicks up 38% and CPC down 8%. Underneath the average, category CPCs on Amazon ranged from $0.55 in clothing, shoes and jewelry to $2.18 in health and household in Q2 2026, according to Pacvue's mid-year report as summarised by beBOLD Digital and logged in Osmos's benchmarks. Skai also records Prime Day 2026 blended CPC at a record-low $1.63.

ROAS equals conversion rate multiplied by average order value, divided by CPC, which makes CPC the input a retailer's auction design controls most directly. Take an illustrative campaign at a $1.00 CPC, a 10% conversion rate and a $50 basket: it returns 5.0x. Lift CPC 11% with conversion and basket unchanged and the same campaign returns 4.5x; let the basket slip as well and the fall compounds. A retailer that lets its auction clear ever higher without adding conversion (relevance ranking, better placements) or basket (bundles, multi-format packages) funds its own yield out of the advertiser's ROAS, and the brand sees the lower return at renewal.

Amazon Ads deep dive

Amazon Ads has compounded scale without giving up much efficiency: Teikametrics' full-year 2025 data, tracked in Osmos's ROAS benchmarks, has its ROAS down only 3.9% while CPC rose 4.6%, and Amazon's advertising services revenue reached $19.8 billion in Q2 2026, up 26% year over year (Amazon, Jul 2026). Sponsored Brands spend rose only 3% and its clicks fell 10% year over year in Q1 2026, while the DSP grew at its quickest pace since Q3 2021 (Marketing Brew, Apr 2026). By Q2 the DSP's clicks had more than doubled with CPC down 35% (Skai, Jul 2026), which is where Amazon's upper-funnel budgets are compounding.

Amazon also reports early numbers for conversational placements: shoppers who click a Sponsored Prompt convert 48% more often and spend 21% more on average, by Amazon's own measure (Amazon, Jul 2026). Amazon has published no method behind these self-reported lifts, so Osmos tracks them as a format signal only. Amazon's ROAS comes from its own traffic, catalog depth and measurement stack, so a retailer using Amazon as a reference should compare the direction of Amazon's numbers with the direction of its own.

Walmart Connect and the scaled second tier

Walmart Connect is the fastest-growing major network in 2026: Walmart Connect in the US grew 43% excluding VIZIO and Walmart's global advertising business grew 38% in the quarter ended July 2026, figures Osmos tracks from Walmart's Q2 FY27 release (Aug 2026). After a 2025 in which advertiser demand outran efficient inventory, its Sponsored Products clicks rose 57% against spend up 62% in Q1 2026 (Marketing Brew, Apr 2026), so clicks were catching up with spend. Walmart display also took 39% of ad spend among advertisers buying both display and search in Q1 2026, up from 35% in Q4 2025, a sign that the second tier is winning upper-funnel budgets as well as search.

Below the two leaders, the other large networks publish advertising revenue but no dated platform-wide ROAS that Osmos's benchmarks can verify for 2026. A retailer benchmarking against them should use the category-matched figures in the industry table and its own auction data.

What does Criteo's research show about retail media ROAS?

Criteo's research shows how much measured ROAS depends on what attribution counts: campaigns that added offline attribution saw an average ROAS increase of 42% compared with online-only attribution, across more than 44,000 retail media campaigns (Criteo, "Beyond ROAS", 2024 data, page updated November 2025), a finding tracked in Osmos's ROAS benchmarks. The same sales, credited across online and store transactions, produce a much higher figure than an online-only report. For a retailer with stores, that is the strongest case for closing the loop on in-store transactions before quoting ROAS to a brand.

How does ROAS differ by ad format?

ROAS differs by ad format because each format sits at a different point in the funnel and clears at a different price, and Osmos's ROAS benchmarks for 2026 show formats moving in different directions on the same network in the same quarter.

Osmos's 2026 format signals, ordered from onsite product ads at the bottom of the funnel to offsite and in-store, from Tinuiti, Marketing Brew, Pacvue and Skai, with the Osmos app that runs each format:

The platform assembles a product ad from the catalog image, title, price and rating, so the listing is the only performance lever for that format. Sponsored brand, display, video and offsite units each take an asset the brand uploads and controls. "Native" describes how a placement renders, and it applies across formats: product ads and custom formats are built to read as part of the page, and the Osmos guide to native advertising benchmarks covers how those placements are measured.

A retailer can build its rate card from this table. Product ads bring volume and the highest direct ROAS, display and video bring the budgets brands renew on measured lift, and offsite reaches demand the onsite shelf cannot hold. Osmos onsite retail media runs every onsite format from one ad server, so a brand's blended return is measured across all of its formats in one system.

What is the average ROAS by industry in 2026?

The average ROAS by industry in 2026 runs from $3.13 in beauty and personal care to $10.19 in electronics, the two category figures in Pacvue's Q2 2026 data for Amazon Sponsored Products, as summarised by beBOLD Digital and tracked in Osmos's ROAS benchmarks (Aug 2026). Those figures describe Pacvue's own Amazon client base. Health and household and clothing carry dated 2026 cost data, and grocery, home and garden, pet, and toys and games have no dated 2026 ROAS multiple that Osmos's benchmarks can verify.

Osmos's 2026 ROAS and cost data by industry, ordered from categories with a dated ROAS multiple to categories with cost data only, then none, from Pacvue's mid-year report (as summarised by beBOLD Digital) and Pacvue's 2026 Beauty and Personal Care Benchmark Report:

Basket size, purchase frequency and CPC explain most of the distance between $3.13 and $10.19. A higher-priced item repays more clicks per sale, which favors categories such as electronics, and a smaller beauty basket needs a higher conversion rate to reach the same multiple. A category with a modest multiple can still be the most profitable inventory a grocery network has, because frequent, low-priced purchases fill the auction every week.

In beauty, Pacvue's category report records spend up 17.5% year over year while CPC rose 10.6% and ROAS fell 5.7%, a combination Osmos's benchmarks read as brands arriving faster than the inventory, with higher click costs pulling ROAS down. A beauty network that adds relevant placements, and holds the price of the slots it already has, gives brands the return the category average is losing.

Treat the electronics figure with care: Pacvue's Q2 2026 figure is the only dated 2026 source for the category, it comes from one vendor's client base, and older category tables placed electronics near the bottom on the view that long consideration cycles depress conversion. The dated 2026 figure is the one to plan against, and a retailer with a large electronics category should benchmark against its own auction data before quoting either.

The hard signal on apparel is price: clothing, shoes and jewelry is the cheapest Amazon category to click on, at $0.55, and a fashion network earns a premium over that price through relevance, which is what filter-matched serving changes. No source in Osmos's 2026 benchmarks publishes an apparel ROAS figure by platform, so the question of which platform has the highest apparel ROAS has no dated answer.

Ecommerce ROAS benchmarks from the open web are not directly comparable with these figures. A retail media ad is served on the retailer's own property, close to the purchase, and credited against the retailer's own transactions, while open-web ecommerce ROAS relies on cookies, pixels and longer paths. Retail media figures therefore read higher on the same spend, and a retailer comparing the two should say so before a brand does.

How can I compare my ROAS to ecommerce ad benchmarks?

Compare your ROAS to ecommerce ad benchmarks only after putting both numbers on the same basis: same category, same ad type, same attribution window and the same rules for what counts as a sale.

1. Match the category. Compare a beauty campaign with a beauty benchmark, because the category spread on one platform runs from $3.13 to $10.19.
2. Match the platform and ad type. A product ad benchmark from Amazon Sponsored Products says little about a display or video campaign, which is judged on lift.
3. Match the attribution window and view-through rules. The IAB and Media Rating Council (MRC) note that "typical attribution windows for digital measurement are often 3, 7, 14, 28, or 30 days," and that "these windows must be disclosed before campaign execution and measurement and as part of disclosures corresponding to reported results" (IAB/MRC Retail Media Measurement Guidelines, January 2024).
4. Separate online, in-store and halo sales. The same guidelines say reporting "should differentiate online sales, in-store sales, and total sales," and Criteo's 42% offline uplift shows how far the split moves the figure.
5. Check the incremental return. Report iROAS next to ROAS, because a benchmark from a vendor's client base measures attributed sales, and only a holdout shows which of those sales the ad created.

A brand's open-web ecommerce ROAS usually rests on a different window and on cookie-based matching, so the honest comparison is on direction and incrementality. The Retail Media Attribution and Measurement playbook sets out window and view-through rules in depth.

How do you measure ROAS in retail media campaigns, and across multiple networks?

You measure ROAS in a retail media campaign by dividing the sales attributed to the ad within a disclosed lookback window by the campaign's cost, and you measure it across networks by putting every network on the same window, sales definition and incrementality test before ranking them.

1. Fix the window and the sale. Set the lookback window, say whether view-through counts, and define an attributed sale at SKU level. The IAB/MRC guidelines define attribution as "the practice of assigning credit for consumer actions, like sales or visits, to specific marketing efforts."
2. Close the loop. Match ad exposure to the retailer's own online and in-store transactions, deterministically where shoppers are logged in, and report online, in-store and total sales separately.
3. Separate branded from unbranded. Branded keywords lift ROAS on every platform by different amounts, so a blended figure hides where the return came from.
4. Hold out a control group. Run an incrementality test on at least one campaign per network, because last-touch and incremental ROAS often disagree about which channel is working.

Across multiple networks, every network sets its own attribution window, view-through rules and definition of an attributed conversion. Comparing a 14-day view-through figure with a 7-day click-only figure ranks the two networks' measurement policies. The test for any display ad platform that claims transparent measurement and ROI is whether it discloses its window, its halo rules and its deterministic or probabilistic matching up front, as the IAB/MRC guidelines require for halo ROAS.

NIQ research from June 2026 found 67% of chief marketing officers plan to increase retail media investment in 2026, but only 53% believe their retail media networks provide adequate measurement and attribution (NIQ, Jul 2026). NIQ's own description of the metric explains the gap: ROAS "measures efficiency, not causality: it tells a brand where its budget went, not what that budget actually created." Osmos Measurement measures every campaign on a retailer's network the same way, and the closed-loop mechanics are set out in closed-loop attribution and ROAS.

For a retailer, the same procedure answers how to measure retail media effectiveness and ROI on its own side: advertiser ROAS and iROAS show the value brands receive, and GMV monetization rate and fill rate show what the network earns from it.

Which commerce media networks deliver the best ROAS for advertisers?

Amazon Ads and Walmart Connect deliver the most proven advertiser ROAS at scale in 2026, and the networks brands rate highest overall are Amazon, Walmart, Kroger, Instacart and Walgreens, the five that lead the Path to Purchase Institute's 2026 Retailer Media Network Ratings (P2PI, Jan 2026). In that survey, 166 brand and agency professionals scored 21 networks on targeting effectiveness, measurement, return on investment (ROI), data sharing, sales growth, creative freedom, traffic-driving and self-service. Each leader sits on high purchase intent and closes the loop against its own transaction data, which is why brands rate them highest.

Osmos reads that ranking as a measure of who already has scale. The figure a Head of Retail Media is accountable for is the return the retailer's own network has to deliver to keep brand budgets renewing, and a smaller network with clean measurement and relevant placements can beat a larger one for a brand in its categories.

Do industry ROAS rankings change the network ranking?

Industry rankings change how a network ranking should be read, because a $3.13 to $10.19 category spread on a single platform in a single quarter, as tracked in Osmos's ROAS benchmarks from Pacvue's Q2 2026 data, means a network's headline return is partly a product of its category mix. A grocery-heavy network carries high-frequency, lower-multiple categories, and a general-merchandise marketplace carries higher-ticket ones. A retailer benchmarking its own network should compare against the category-matched figure, and should expect a brand in a $2.18-CPC category to judge the network's price on that basis.

What benchmarks exist for CTV advertising ROAS?

The honest benchmark for connected TV is incremental sales lift against a holdout, and no dated 2026 source in Osmos's ROAS benchmarks publishes a CTV ROAS by industry. A product ad's ROAS is measured against transactions on the same property inside a short attribution window. A CTV exposure and the purchase it drives are usually separated by days and devices, so a ROAS measured the same way understates CTV by design. A network quoting a headline CTV ROAS should state the attribution window and whether the measurement was incremental or last-touch, and a retailer offering CTV should lead with the lift study.

Which traffic sources do ROAS-focused advertisers trust?

ROAS-focused advertisers trust traffic they can tie to a sale: on-property search and category placements on a retailer or marketplace, checkout and post-purchase placements where the transaction is recorded, and offsite audiences built from the retailer's first-party data and reported closed-loop. Checkout monetization platforms and agencies are judged the same way. The benchmark reports buyers quote come from buy-side tools and analysts, since agencies do not publish comparable ROAS league tables, which is why Osmos's compilation names each source and its client base. For a retailer, commerce media traffic earns trust when every placement reports to the same sale.

Which ROAS optimization tools offer transparent pricing?

Pricing transparency comes down to whether the vendor takes a percentage of the media it optimizes. A tool priced as a share of ad spend earns less when the brand spends less, which gives the vendor an incentive that runs against efficiency, and a flat license or per-seat price removes that incentive. A retailer evaluating anything that touches its auction should ask in writing whether the vendor ever takes a position in the auction it reports on, whether clearing prices are visible, and what the effective rate becomes at full volume after the pilot. On the retailer's side, Osmos Ad Packages show real-time availability, rule-based volume discounts and exclusivity tiers at booking, so brands see the price before they commit.

What ROAS does a retailer's own network need to deliver to retain advertiser budgets?

A retailer's network needs to deliver a return that clears the brand's blended return on its existing paid search and paid social, measured on a comparable attribution basis, and prove it with reporting the brand trusts. A brand renews when the network beats the alternative use of that budget, so the brand's other channels set the benchmark. The network-side scorecard that sits behind it, including fill rate and advertiser retention, is in Retail Media Benchmarks 2026: What Success Looks Like.

A retailer controls the inputs to that return before a brand ever sees a ROAS figure: how relevant each ad is to what the shopper is looking at, how much of its inventory it fills, and how fast it serves an ad at peak traffic. Osmos replaced the in-house ad server at one of India's largest fashion and lifestyle e-commerce marketplaces, and during a roughly three-week flagship sale in 2026 Osmos's ad infrastructure delivered a 2.5x fill rate increase over the previous in-house platform, a 40% CTR lift against the non-sale period and under 50ms p95 ad serving latency held through peak load, while the retailer reached 3.5x its average daily sales during its biggest annual sale. The CTR lift came as ad serving matched what shoppers had actually filtered for. Those are supply-side results, measured on fill, click-through and latency, and they describe the conditions under which a brand's return holds when traffic peaks. For the retailer's own business case the network-level return matters as much as the brand's, and margin is the number to plan on.

What are the key KPIs in commerce media measurement?

The key KPIs in commerce media measurement are attributed ROAS, incremental ROAS, ACoS, new-to-brand rate, conversion rate, CPC and, for the retailer itself, GMV monetization rate, fill rate and advertiser retention. Seven in ten brands meet or exceed their retail media goals (Skai, State of Retail Media 2026, Jan 2026), and the industry now decides whether a network is worth renewing on more than ROAS alone. Jason Wescott, Global Head of Commerce Solutions for WPP Media, described the shift in Skai's 2026 measurement report: "The overreliance on ROAS as the benchmark of value is over" (Skai, The 2026 State of Retail Media Measurement and Incrementality, Feb 2026).

Osmos's KPI benchmarks beyond ROAS, ordered from the return metrics a brand reports to the price and proof behind them, from Teikametrics, Skai, Pacvue and the IAB/MRC guidelines:

For a retailer building or scaling a network, the KPI hierarchy runs: GMV monetization rate (revenue efficiency), advertiser ROAS and iROAS (the value proposition to brands), fill rate (inventory utilization) and advertiser retention (sustainability). A brand that acquires customers through a retailer's network at a lower blended cost than paid social keeps renewing, which ties advertiser retention to the customer acquisition cost argument in How Retail Media Offsets Customer Acquisition Costs.

Osmos Measurement builds these KPIs into the campaign. Brand Lift Studies run exposed-versus-control surveys that track recall, awareness, consideration and purchase intent, so brands see the upper-funnel impact ROAS misses. Osmos Incrementality, in early access, assigns eligible shoppers to an exposed group and a holdout on a 95/5 split by default, configurable per campaign, and reports incremental conversions, revenue and ROAS per campaign, with a minimum measurement window of eight weeks. It turns on at campaign launch, so a retailer's first incrementality proof does not need a separate study or a second contract.

How does a retailer improve the ROAS its network delivers?

A retailer improves the ROAS its network delivers through relevance, a forecast before the spend, format breadth and complete attribution.

1. Relevance before price. Conversion is the input a network can raise without costing the brand more. Osmos Product Ads auto-select products from the catalog based on 50+ behavioral signals, so a brand launches with relevant SKUs, and floor CPCs adjust to demand on every slot. The Osmos guides to AI-powered ad serving and behavioral targeting and contextual vs behavioral targeting cover how relevance is built.
2. A forecast before the spend. The Osmos Media Forecaster recalculates spend, impressions, orders, revenue and ROAS in real time as a brand changes campaign settings, so a brand sees the likely return before it commits. This is the closest thing to a ROAS prediction platform inside a retailer's own console.
3. Format breadth. Among organizations running full-funnel retail media strategies, 48% cite improved performance and ROAS as a benefit (Skai, Jan 2026). A network that offers only a listing format has no inventory for the budgets moving to display and video. Osmos Ad Packages bundle product ads, display, video, in-store screens and offsite into one unit with a defined flight and price, with promised-versus-delivered reporting on impressions, clicks, video plays, attributed orders and revenue.
4. Attribution completeness and trusted measurement. A network that attributes in-store transactions as well as online ones reports the return its offline-heavy categories actually earn, and one that supplies the holdout read fills the confidence gap most brands report. First-party data provides the targeting layer that makes this attribution deterministic.

How do retailers build a high-ROAS network on Osmos?

Retailers build a high-ROAS network on Osmos by running onsite, offsite and in-store retail media from one system, so every format a brand buys is measured against the same sales. Retailers monetize a retail media network by offering ad placements on the surfaces they own, priced on CPC, CPM or by store and day, with first-party shopper data as the targeting layer and measurement that proves sales keeping brands renewing. The benefit for a retailer is a high-margin revenue line on traffic it already has, and the 2026 revenue strategies that work start with onsite product ads and add formats as measurement matures.

Onsite.Osmos onsite retail media runs Product Ads, Product Display Ads, Display Ads, Video Ads, Custom Ads and Non-endemic Ads from one ad server, with dynamic floor CPCs, brand distancing that keeps same-brand products from appearing side by side, and full-funnel attribution back to the SKU. All targeting runs on first-party shopper data.

Offsite.Osmos offsite retail media extends that first-party data off the retailer's properties. Its White Labeled DSP runs the retailer's audience network across Meta, Google, TikTok and The Trade Desk, and Guaranteed Outcomes runs tail and torso advertisers across Meta, Google and TikTok, managed to outcomes, so the retailer earns offsite ad revenue without adding ad-ops headcount. Osmos Audience Exchange and Bring Your Own Traffic are in early access.

In-store.Osmos in-store retail media turns the store estate into bookable inventory through a Retail Media CMS, a Booking Calendar, Physical Media for analog placements and Programmatic Demand, booked and priced by store, day and campaign.

Marketplaces follow the same path with more onsite surfaces and fewer stores, as marketplaces that build their own ad platform show, and the media line adds margin on top of the merchandise business. To see how the Osmos platform would run your network, get a demo or explore the Osmos platform.

Frequently Asked Questions

How does Amazon ROAS compare to Walmart Connect in 2026?

Amazon held efficiency better through 2025, with ROAS down 3.9% against about 18% for Walmart in Teikametrics' data, and Walmart Connect is growing faster in 2026. Because the two networks use different attribution windows and category mixes, compare the direction of each network's numbers, quarter by quarter.

What do CommerceIQ, Skai, Pacvue and Perpetua benchmark reports tell a retailer?

They are buy-side platforms, and their benchmarks describe what their own brand clients earned on the major networks, which is useful precisely because it is the figure a brand will bring to a renewal. Skai and Pacvue publish dated quarterly reports, and Pacvue's category figures cover its Amazon Sponsored Products clients. A retailer should read all four as a record of advertiser expectations and read the networks' own earnings releases as the network-side record.

How does retail media ROAS compare to traditional advertising ROI?

The clearest 2026 signal is the direction of budgets: 52% of advertisers are reallocating display spend from open-web demand-side platforms to retail media DSPs, with 57% of them naming closed-loop attribution as the reason (Skai, State of Retail Media 2026, Jan 2026). No dated 2026 source publishes a like-for-like ROI figure, because traditional media is measured on reach and modeled sales while retail media is measured on attributed transactions.

What should a retailer quote to a brand in a category with no published benchmark?

Quote the network's own category-matched auction data: the conversion rate and CPC the category clears at on the network, set against the CPC the brand already pays in that category elsewhere. Offer an incrementality read on the first flight so the brand's number comes from its own campaign. A retailer can use a figure from a comparable category on another platform as a reference point.

Can a retail media network offer ROAS guarantees or performance-based pricing?

A network can offer an outcome-defined product even where a ROAS guarantee is impractical: an Osmos Ad Package commits to delivery and proves it package by package, and Osmos Guaranteed Outcomes manages tail and torso advertisers to outcomes across Meta, Google and TikTok. Contractual ROAS guarantees remain rare, because attribution methods vary too widely for a binding cross-platform commitment.

How is GMV monetization rate calculated and tracked?

Divide retail media revenue by GMV, or by gross transaction value for a marketplace, over the same period, and track it monthly by category as well as for the network as a whole. Categories where brands already pay the highest CPC elsewhere, such as health and household, are the natural place to test pricing first.

What is the future of retail media ROAS measurement?

Incrementality is moving from a bespoke study to a setting inside the campaign, and in-store measurement is maturing toward verified delivery, per the IAB framework for in-store media measurement published in December 2025. Expect renewal conversations by 2027 to turn on incremental ROAS, new-to-brand customers and organic halo, with reported ROAS as the tie-breaker. If a network does not offer a holdout group, brands run their own incrementality tests to check the reported return.

Which 2026 retail media benchmark reports are dated, and how often do they update?

Quarterly: Skai's digital advertising trends reports (Q1 2026 in April, Q2 2026 in July, with ROAS in the Q1 edition only), Pacvue's commerce media benchmark (mid-year edition July 2026, category reports in August) and the earnings releases from Amazon (July) and Walmart (August). Annual or periodic: the Teikametrics Amazon versus Walmart benchmark and Tinuiti's retail media outlook, both March 2026, NIQ's incrementality research (July 2026), and forecasts from eMarketer (January and May 2026), WARC (August 2026) and Forrester (October 2025). Q3 2026 releases land from late October, so set targets from benchmarks published within the last four quarters.

Related Deep Dives

• Retail Media Benchmarks 2026: What Success Looks Like: the scorecard a network is judged on once ROAS is in hand.
• Retail Media Ads: Sponsored Formats 2026: the unit types, how each is priced and where each sits in the funnel.
• Native Advertising Retail Media Benchmarks 2026: benchmarks for placements that render as part of the page.
• Onsite vs Offsite Retail Media Ad Formats Framework 2026: when a format belongs on the retailer's own properties and when it belongs off them.
• How Retail Media Offsets Customer Acquisition Costs: the acquisition-cost arithmetic that keeps brands renewing.
• How Retailers Use Retail Media to Boost Margins: how advertising revenue changes a retailer's margin mix.

Sources

1. Skai. "Skai's Q1 2026 Digital Advertising Trends Report: Key Takeaways for the Year Ahead." April 2026. skai.io
2. Skai. "Ad Spend Rises Across Every Channel in Q2 2026 as AI Reshapes Commerce Media, Skai Data Shows." July 2026. skai.io
3. Skai. "What is iROAS and Why It Matters for Retail Media Performance." Updated November 2025. skai.io
4. Skai, with Stratably. "State of Retail Media 2026." January 2026. skai.io
5. Skai. "Retail Media Reaches Inflection Point as GenAI and Omnichannel Strategies Reshape Advertising Spend." January 2026. skai.io
6. Skai. "The 2026 State of Retail Media Measurement and Incrementality." February 2026. skai.io
7. beBOLD Digital. "Amazon Q2 2026 retail media benchmarks" (summary of Pacvue's 2026 Mid-Year Commerce Media and Summer Sales Benchmark Report). August 2026. bebolddigital.com
8. Pacvue. "2026 Mid-Year Commerce Media and Summer Sales Benchmark Report." July 2026. pacvue.com
9. Pacvue. "2026 Beauty and Personal Care Benchmark Report." August 2026. pacvue.com
10. Criteo. "Beyond ROAS: A broader approach to measuring retail media success." May 2024, updated November 2025. criteo.com
11. Amazon.com, Inc. "Amazon.com Announces Second Quarter Results." July 2026. ir.aboutamazon.com
12. Amazon. "Q2 earnings: CEO Andy Jassy on what's driving Amazon Ads growth." July 2026. aboutamazon.com
13. Walmart. "Walmart Releases Q2 FY27 Earnings." August 2026. corporate.walmart.com
14. IAB and Media Rating Council. "Retail Media Measurement Guidelines." January 2024. iab.com
15. IAB. "A Viable Framework for Maturing In-Store Media Measurement." December 2025. iab.com
16. NIQ. "How Brands Can Measure Incrementality and Prove Retail Media Impact." July 2026. nielseniq.com
17. eMarketer. "FAQ on retail media networks: How marketers should allocate budgets in 2026." January 2026. emarketer.com
18. eMarketer. "Retail Media Ad Spending Forecast H1 2026." May 2026. emarketer.com
19. WARC, via The Desk. "WARC: Retail media ad market to top 200 billion as growth slows." August 2026. thedesk.net
20. Forrester. "Global Retail Media Spend To Top $300 Billion By 2030." October 2025. forrester.com
21. Teikametrics. "2026 Amazon vs Walmart Benchmark Report." March 2026. teikametrics.com
22. Tinuiti. "Retail Media Trends and Outlook for 2026." March 2026. tinuiti.com
23. Marketing Brew. "Ad spend on YouTube, Google Search, Amazon Sponsored Products is growing: report." April 2026. marketingbrew.com
24. Path to Purchase Institute. "2026 Retailer Media Network Ratings: Brand, Agency Execs Grade Walmart, Amazon, Kroger, Others." January 2026. p2pi.com