Uber Advertising today unveiled a measurement product that lets restaurant advertisers in the United States trace what Sponsored Listings on Uber Eats do for sales at physical restaurant locations, using bank card transaction data supplied by Affinity Solutions. According to Uber Advertising, an accompanying report puts the all-channel sales increase at 27%, against 40% on the delivery app itself.

In Short

Uber today showed restaurants a way to check whether ads on its delivery app also lead to more spending at the restaurant's own physical locations, using anonymized bank card data from a company called Affinity Solutions. Delivery platforms normally see only the orders placed on their own app, so what an ad does to a later walk-in visit has mostly gone uncounted. What changes is that US restaurant advertisers can now get a number for in-restaurant sales, although the figures come from Uber and Affinity rather than an outside auditor.

A card-data layer on top of Uber's own experiments

The product is called Sponsored Listings Cross-Channel Incrementality, and Uber Advertising presents it as an extension of a method already in use rather than a new one. Uber Eats opened Sponsored Listings to US restaurants in August 2020as a pay-per-click format, and the company's existing Sponsored Listings Incrementality measurement, according to Uber Advertising, uses experimentation to determine the causal effect of those ads on sales inside Uber Eats.

The new layer keeps that test-and-control design and adds Affinity Solutions' transaction records. Uber Advertising describes the source as deterministic, bank-direct credit and debit card data, and says the findings it receives are deidentified and aggregated. The stated scope is the United States.

Jake Kastner, Global Head of Measurement Strategy at Uber Advertising, said the work began with a narrower question: whether a sale on Uber Eats cannibalizes an offline order, "a question that gets to the heart of how we think about incrementality." Kalyan Lanka, Chief Product Officer at Affinity Solutions, said "Marketers shouldn't have to treat what happens off-platform as a blind spot."

Both companies supply their own scale figures. Uber Advertising says it connects brands with more than 200M monthly active users across Uber and Uber Eats. Affinity Solutions says it has permissioned purchase data from over 100 million US and UK cardholders, representing 86 billion transactions, and that its proprietary AI technology, Comet, converts those transactions into insights. The panel therefore spans two countries, whereas the product described today is limited to restaurant locations in one.

What the report found

Alongside the product, the two companies published a report titled Beyond Delivery: Driving Holistic Restaurant Growth with Uber Advertising. According to Uber Advertising, it examines the cross-channel and longer-term effect of Sponsored Listings on restaurant growth, reporting sales lift across several channels and periods. The figures below are vendor-supplied, and the report itself was not among the materials available for this article.

MeasureResultPeriod
Sales increase on Uber Eats40%During the campaign
Sales increase across measured channels27%During the campaign
Sales increase at brick-and-mortar locations5%During the campaign
Average sales lift on Uber Eats8%Following six months
Average sales lift at brick-and-mortar locationsApproximately 5%Following six months
Sales lift for new-to-brand eaters15%Not stated

The during-campaign figures refer to eaters shown Sponsored Listings, as opposed to a restaurant's total sales, and the release does not restate the population behind the six-month averages. The 15% figure for new-to-brand eaters appears in the materials accompanying the announcement rather than in the release text, and Uber Advertising's materials add that the largest off-platform increases occurred among infrequent and first-time customers of a brand.

Read side by side, the numbers describe two different decays. The effect on the app falls to about a fifth of its in-campaign size once campaigns end, from 40% to an 8% average. At physical locations, by contrast, the post-campaign average of roughly 5% matches the in-campaign result. Whether the two periods share a baseline is not specified, so the comparison remains approximate.

Two of the figures are also not new. Uber Advertising published a 27% incremental sales lift across all channels and a 5% lift in brick-and-mortar sales on June 22, attributing them then to a study run with Affinity Solutions in May 2026 across 12 campaigns. Whether today's report rests on those same 12 campaigns or on a larger set is not stated. The same gap applies to the 40%, 8% and 15% results: the materials give no count of campaigns, restaurants or eaters behind them.

Where the method leaves questions open

A sales lift study compares purchases by people who saw a campaign with purchases by a control group that did not, and credits the difference to the advertising. The strength of the result depends on how the control group is built. A randomized holdout, where eligible users are split before delivery, sits at the top of most evidence hierarchies. A matched control built after the fact sits lower.

The announcement does not say which applies. It also leaves unstated how an app user is linked to a cardholder, what share of restaurant spending the card panel can see, how walk-in purchases are told apart from the app's own orders in card records, and whether the reported differences carry confidence intervals. Each of those determines how far a 5% figure can be trusted.

The question has a standard against which it can be placed. IAB and IAB Europe published guidelines for incremental measurement in commerce media on November 3, 2025, sorting methods into four tiers of evidence; randomized tests, holdouts and matched markets rank highest, and results reported by the platform that sells the ads rank lowest. Where Uber's design falls in that ladder turns on assignment details that the release does not provide. The card data does come from a party other than the seller, which is a separate matter from the experiment design.

Delivery apps building their own proof

Delivery apps belong to what the industry calls commerce media, advertising sold by companies that hold shopper transaction data. Their recurring weakness is the boundary of the data: the seller sees the order it processed and little else. Uber Advertising has been addressing that boundary from several directions during the past year.

In December 2025 it opened Uber Intelligence on LiveRamp's clean room, letting brands combine their customer data with Uber's consented signals. On June 2, 2026 it opened Ads APIs to point-of-sale providers, aggregators and consumer goods brands. A competitor moved the same week: DoorDash, on June 4, added a LiveRamp clean room so advertiser and DoorDash first-party data can be matched for overlap and incrementality analysis. Today's product differs in drawing on bank card transactions rather than a data collaboration environment.

Affinity Solutions, for its part, has been supplying card data to several measurement products. Cint added Affinity transaction data as the sales lift signal in its Lucid Measurement dashboard on June 17, 2026. StackAdapt followed on July 30 with a report tied to Affinity data covering more than 150 million cards and $4 trillion in spending. The units differ between that description (cards) and Affinity's own boilerplate in today's release (cardholders), so the two figures are not directly comparable.

Affinity also sits in the middle of a measurement argument. Its May 2026 survey of marketers found 91% saying platform-reported results are overstated to some degree. Uber's new product pairs a platform-run experiment with an outside data source, which addresses part of that concern and leaves another part, the experiment's design, with the platform.

Why this matters for the marketing community

Measurement in this corner of advertising, in Uber Advertising's own description, often stops at the platform boundary. A figure for physical locations, if it holds up, changes the arithmetic of what a Sponsored Listings budget is worth, particularly for chains whose sales happen mostly outside the app. The announcement does not give the channel mix behind the blended 27%, so the weight of the 5% in-restaurant effect within a restaurant's total revenue cannot be derived from it.

The duration question may prove as consequential as the channel question. Uber Advertising's materials argue that measurement which stops when a campaign ends can understate results; the six-month figures are offered as support. That is an argument from the vendor's own data, and the 8% and roughly 5% averages arrive without the campaign counts or intervals that would let a buyer test them.

There is also the matter of precedent. Delivery and mobility platforms, payment companies and card-data suppliers are converging on the same premise, that a platform's ad product is only as credible as its proof beyond its own walls. Will buyers accept proof that the seller commissions, even when part of the data comes from elsewhere? Today's report adds one more data point to that debate, and the methodological detail still missing will decide how much weight it carries.

Timeline

Summary

Who: Uber Advertising, the advertising division of Uber, and Affinity Solutions, a consumer purchase data company; the audience is restaurant advertisers in the United States.

What: Sponsored Listings Cross-Channel Incrementality, a measurement product that adds Affinity Solutions' deidentified, aggregated bank card transaction data to Uber Advertising's test-and-control experiments, plus a report, Beyond Delivery, reporting a 40% sales increase on Uber Eats, 27% across measured channels and 5% at brick-and-mortar locations during campaigns, and an 8% and roughly 5% average effect over the following six months.

When: Today, October 5, 2026.

Where: Restaurant advertisers using Sponsored Listings on Uber Eats, with the cross-channel view covering physical restaurant locations in the United States.

Why: Commerce media measurement usually stops at the platform boundary, which leaves advertisers without a view of the sales a campaign drives elsewhere or after it ends; Uber Advertising and Affinity Solutions present the product as a way to close that gap.