Circana and Best Buy Ads on October 7, 2026 published a marketing mix modeling study claiming that nearly 16% of the media-driven sales associated with Best Buy's retail media business occurred at competing retailers, that its return on ad spend ran 79% above all other media, and that it accounted for nearly 10% of measured brand equity value. The release names no brands, no time period, no spend levels and no sample size.

In Short

Best Buy sells ads on its website, app and stores, and a research firm called Circana built statistical models suggesting those ads also push people to buy the same products at other shops. That matters to the electronics and appliance brands deciding whether to put money into retailer ads or into national TV and online campaigns. What changes is the argument: Best Buy can now say its ads do more than move sales from one checkout to another, although the study behind that claim has not been published in detail.

What Circana and Best Buy Ads claim

The announcement was issued from Chicago and Minneapolis on October 7, 2026, and appears on Circana's website under the byline of Lindsay Pullins, the company's senior vice president of Retail Media, who is also quoted in it. Circana described the work as a "first-of-its-kind marketing mix modeling (MMM) study" and as "new syndicated marketing mix modeling research" - language that signals a study spanning several advertisers rather than a model built for a single client.

According to Circana, the study "analyzed leading brands across consumer electronics and appliances" and "evaluated Best Buy Ads against national media investments to determine how retail media contributes to sales, category demand, and long-term brand value." Three headline findings follow, each attributed to "Circana MMM analysis":

  • Return on ad spend. "ROAS for Best Buy Ads was 79% higher than all other media based on total U.S. Best Buy sales."
  • Brand equity. "Best Buy Ads contributed nearly 10% of measured brand equity value."
  • Cross-retailer sales. "Nearly 16% of media-driven sales associated with Best Buy Ads occurred at competing retailers."

All three figures are company-supplied. Neither company has published a methodology document, a report or a technical appendix alongside the release, and none is linked from the announcement page.

Two of the three findings carry an interpretation that the release explicitly assigns to Best Buy rather than to Circana. After the brand equity figure, the text reads: "Best Buy believes this demonstrates retail media's ability to build long-term brand strength alongside immediate business results." After the cross-retailer figure: "Best Buy believes this indicates that its retail media influence extends beyond Best Buy's ecosystem and contributes to category demand." The distinction matters. The numbers come from the model; the conclusions drawn from them are presented as the retailer's view.

The release then states more broadly that "the research documents a measurable halo effect," that Best Buy Ads "influenced purchases that occurred beyond Best Buy's own channels," and that retail media "can generate category demand across the broader marketplace rather than simply shifting where a purchase happens."

What the two executives said

Pullins framed the study as filling a measurement gap. "This study gives advertisers something they have long needed: a complete, independent view of what retail media actually delivers," she said, according to Circana. "The results are clear. Best Buy Ads drives measurable incremental sales, influences purchases across the wider market, and builds lasting brand equity. For marketers deciding where the next dollar should go, that full-funnel evidence is exactly the kind of clarity that leads to better investment decisions."

Lisa Valentino, president of Best Buy Ads, pointed at the limits of the measurement retailers usually provide. "For years, retail media measurement has focused primarily on what happens inside a retailer's own walls," she said. "This research shows the impact is much broader. Best Buy Ads helps brands drive immediate sales, strengthen customer consideration, and influence purchases across the marketplace. As advertisers seek greater accountability across both performance and brand-building investments, it's critical to understand the full value retail media can deliver."

What the release does not disclose

The word "independent" appears in both the release body and Pullins' quote. The study was, however, published jointly with the company whose advertising it evaluates, and the release does not say who commissioned or paid for it. Circana sells measurement services to retailers and brands; Best Buy sells the advertising being measured. That arrangement is common in the industry, and it does not by itself invalidate the results, but it sits uneasily alongside the description of an independent view.

A long list of basic parameters is absent from the document:

  • Brands. "Leading brands across consumer electronics and appliances" is the only description. The number of brands, and whether they were Best Buy Ads clients at the time, is not stated.
  • Time period. No start or end date for the modelled data appears in the release.
  • Spend. There is no figure for what the brands spent on Best Buy Ads or on the national media they were compared with.
  • Comparison set. "All other media" and "national media investments" are not broken down into channels such as linear TV, connected TV, search, social or other retailers' ad networks.
  • Channel mix inside Best Buy Ads. The release does not say whether the measured activity was sponsored listings on BestBuy.com, offsite display, connected TV, in-store screens or a combination. That matters because, as other research shows, the share of sales that leaks to other retailers varies sharply by format.
  • Brand equity definition. "Measured brand equity value" is not defined. Nothing in the release explains how equity was measured or what the remaining 90% consists of.
  • Model specification. The number of models, the modelling technique, the confidence intervals around each finding and whether results were validated against experiments are not given.

The ROAS base

The wording of the 79% figure deserves a close reading. ROAS for Best Buy Ads was higher "based on total U.S. Best Buy sales." Taken literally, the comparison counts returns only where they landed at Best Buy. If national media investments - a TV campaign for a television brand, for example - drive purchases at every retailer that stocks the product, then measuring their return solely on Best Buy's tills would capture only a fraction of what they produced. Ads sold by Best Buy itself, aimed at Best Buy shoppers, would by design concentrate more of their effect on that same base.

This is an inference from the wording, not something the release addresses, and the full model may handle it differently. But the same release makes the case that Best Buy Ads generates nearly 16% of its media-driven sales elsewhere. Read together, the two claims use different scopes: the ROAS advantage is measured inside Best Buy, while the halo is measured across the market. Whether the 79% gap would narrow, widen or hold if every channel were credited with sales across all retailers is not disclosed.

The halo figure in context

Cross-retailer effects have become one of the more contested numbers in retail media. In June 2026, measurement company Incremental published a white paper covering more than 150,000 campaigns and $350 million in spend across four unnamed retail media networks, which found that siloed attribution misses an estimated 36% to 53% of total retail media impact. Its breakdown varied sharply by format. Search ads kept 93% of their impact at the advertising retailer, with 6% appearing elsewhere. On-site DSP placements sent 29% of their impact to other retailers, and off-site DSP sent 52%.

Against that spread, Best Buy's "nearly 16%" sits between the search figure and the on-site display figure. Without knowing the mix of formats Best Buy Ads ran for the brands in the study, though, the two numbers cannot be compared directly. Incremental's dataset covered CPG brands; Circana's covers consumer electronics and appliances, categories with higher prices, longer consideration windows and a different retail landscape.

Amazon has been pursuing a similar argument through product rather than research. On July 7, 2026, Amazon Ads told advertisers that its Omnichannel Metrics reporting in Amazon DSP now names the specific retailers its ads drive sales to, for US display, audio and video campaigns. That tool uses exposure-based attribution built on the Amazon Shopper Panel. Circana's approach is different in kind: MMM works from aggregated time series of spend and sales rather than from individual exposure records, which means it can estimate effects at competing retailers without needing those retailers' transaction data at user level.

The commercial logic is plain. A retailer that can show its ads lift sales at rival chains can argue that its media budget belongs alongside national brand campaigns, not inside the trade marketing or shopper budget where retail media has traditionally been funded. Valentino's quote makes that pitch directly, as does the release's closing claim that the study "equips advertisers to make more informed decisions across both performance marketing and brand-building objectives."

Why MMM, and why now

The choice of marketing mix modeling is itself part of the story. MMM estimates the contribution of each channel to sales using historical spend and outcome data, and it does not depend on cookies or user-level tracking. Retail media has had an awkward relationship with the method. In April 2026, the IAB published a white paper written with Instacart arguing that legacy measurement is cheating retail media out of its real value. That paper cited a February 2026 study in which 50% of MMM users said commerce media was underrepresented in their models. The IAB position was that MMM is best kept as a portfolio planning tool, with incrementality experiments - control and exposed groups, randomised tests and geographic lift studies - acting as the primary foundation for retail media measurement.

The Circana and Best Buy study takes a different route to a similar destination. Rather than arguing that MMM undercounts retail media, it uses MMM to make the case that retail media outperforms. The release does not say whether any experimental results were used to calibrate the models, which is the approach the IAB recommended and which Google's open-source Meridian model supports through experiment-based priors. Google opened Meridian to all marketers in January 2025.

Circana's own position in MMM has grown considerably. In August 2024, the company agreed to acquire NCSolutions and Nielsen's marketing mix modeling business, folding both into its media team. The announcement page for the Best Buy study also lists a separate Circana post dated September 8, 2026, titled "Circana Advances AI-Powered Marketing Measurement: Introducing Liquid Mix with Google Meridian," indicating that the company now offers an MMM product built on Google's framework. Which modelling engine produced the Best Buy results is not stated.

A comparison with Circana's YouTube study

Circana has produced platform-backed MMM research before, and the earlier example shows how much detail such studies can carry. In May 2026, YouTube distributed a creator marketing playbook whose headline figure - 86% higher incremental long-term ROAS than paid social - came from a Circana study commissioned by Google. That study disclosed its sample: 20 brand-level observations from 40 marketing mix models, covering 10 US CPG brands across home care, personal care, beauty, food and beverage, over 104 weeks in 2023 and 2024. It also disclosed that the ROAS figure covered total Google media rather than YouTube alone.

None of those parameters - brand count, number of models, observation window, or the precise scope of the media being credited - appears in the Best Buy release. Both studies share a structure: a single platform's ads compared against a broad category of "other" media, using long-term brand effects as part of the case. Platform-funded MMM has become a familiar format. Kochava's 2025 work for TikTok, which found 35% higher incremental impact than last-touch attribution, followed a similar pattern of a measurement firm publishing favourable model outputs for a media seller.

Best Buy Ads and its data assets

Best Buy's ad business has been expanding its programmatic and offsite footprint for more than a year. In September 2025, the retailer selected Magnite as its exclusive supply-side platform and curation partner for onsite and offsite inventory. At that time, Best Buy said its media business handled more than two billion customer interactions a year from nearly 80% of the US adult population, worked with more than 600 brands, and could tie 93% of transactions to customer IDs across store, app and online purchases. That 93% figure is the foundation of closed-loop reporting: linking ad exposure to a later purchase inside the retailer's own records.

In April 2026, Best Buy Ads became one of six launch partners in Roku Curate, contributing purchase and browsing data from more than 200 million customers to pre-packaged connected TV buys with closed-loop measurement. Valentino, quoted there as president of Best Buy Ads, said the partnership was about "driving awareness and consideration with clear visibility into downstream impact." In June 2026, Perion won the contract for Best Buy Canada's 308-store retail media network, moving in-store screens from fixed loops to programmatic decisioning.

Each of those moves pushes Best Buy Ads further from sponsored product listings on its own site and toward channels - connected TV, display, in-store screens - where closed-loop data captures only part of the effect. Closed-loop measurement, by construction, sees only Best Buy transactions. A study claiming that nearly 16% of the effect happens outside those transactions is, in that sense, an argument that Best Buy's own reporting undercounts its value. That is the gap Valentino's quote about "a retailer's own walls" describes.

Why it matters for advertisers

The study lands in a market where buyers are both spending more on retail media and asking harder questions about it. At IAB Australia's July 2026 summit, a trade survey showed that 60% of buy-side respondents increased retail media investment over the past year, while nearly half of that investment was reallocated from other channels. At the same event, Metcash warned that its 800-screen in-store network was losing brand dollars over incrementality questions. In Europe, IAB Europe figures put retail media spending at EUR 13.7 billion in 2024, with a forecast of EUR 20.8 billion for 2026.

Retail media networks have competed largely on attributed ROAS, a metric that favours lower-funnel formats sitting close to the point of purchase. The Circana and Best Buy study tries to widen the frame in two directions: outward to sales at other retailers, and forward to brand equity. If accepted by advertisers, that framing would let a retail media network compete for budgets currently held by national TV and digital brand campaigns.

Whether advertisers accept it depends on details the release does not contain. For an electronics or appliance brand reading the findings, three questions are unanswered. First, would the same model show similar cross-retailer effects for its own products and category? Second, what was Best Buy Ads actually compared with, and on which sales base? Third, how sensitive are the results to the model's assumptions about long-term brand effects, a component that tends to be the least observable part of any MMM? Circana and Best Buy may be able to answer all three for individual clients. The public release does not.

There is also a structural point about who produces the evidence. Retailers, platforms and measurement vendors increasingly co-publish studies that support the case for a specific seller's inventory. YouTube's Circana study, TikTok's Kochava study and now Best Buy's Circana study each arrived with a favourable headline number. Independent replication across multiple retailers in the same category - Amazon, Walmart, Target and Best Buy measured with the same model - would tell advertisers far more about where retail media budgets work than any single-network study can.

Timeline

Summary

Who: Circana, the Chicago-based consumer data and measurement company, and Best Buy Ads, the advertising business of US electronics retailer Best Buy. Lindsay Pullins, senior vice president of Retail Media at Circana, and Lisa Valentino, president of Best Buy Ads, are quoted.

What: A syndicated marketing mix modeling study of unnamed consumer electronics and appliance brands claiming Best Buy Ads delivered ROAS 79% higher than all other media (measured on total US Best Buy sales), contributed nearly 10% of measured brand equity value, and saw nearly 16% of its media-driven sales occur at competing retailers. All figures are company-supplied, and the release discloses no brand count, time period, spend, channel breakdown or model specification.

When: The study was announced on October 7, 2026.

Where: The announcement was issued from Chicago and Minneapolis, and the findings refer to US sales.

Why: Best Buy Ads is using the results to argue that retail media works as a brand-building channel with effects beyond its own stores, a case aimed at budgets now held by national media. The study arrives as advertisers raise retail media spending while questioning how its incrementality is measured, and as the industry debates whether MMM undercounts or overstates commerce media.