Refinery89, an ad tech company that sells inventory from a publisher network across Europe and Latin America, said on September 14, 2026 that it had wired Brand Metrics questionnaires into the campaigns it runs, promising buyers a reading of how audiences perceive a brand while the media is still live.

In Short

Refinery89, which sells ad space on websites and apps in Europe and Latin America, now works with Brand Metrics so advertisers can see how people feel about their brand while a campaign is still running. That matters to brands and agencies, which often struggle to show what awareness advertising achieved until long after the money has been spent. Short surveys shown to readers now report awareness, consideration, preference and purchase intent during the campaign, although prices, minimum campaign sizes and the survey design have not been published.

What the two companies agreed

The deal, which Refinery89 described as a strategic partnership, attaches Brand Metrics' measurement to a product Refinery89 calls Brandformance. According to Refinery89, advertisers buying through its network of publishers - described by the company as boutique and brand-safe - can now commission Brand Lift Studies as part of that package. The integration has already rolled out, the companies said, and is now being extended across Europe and Latin America.

Much of the operational detail is missing. Neither company disclosed when the first studies ran, which countries are live, how many publishers sit in the network or how many campaigns have been measured so far. Whether a study carries a charge on top of the media cost is not stated. No minimum budget or impression floor is mentioned anywhere in the material.

The geographic framing also shifts from one paragraph to the next. The announcement's headline refers to global brand growth, and its opening paragraph speaks of advertisers and agencies worldwide. The rollout paragraph, and the statement from Brand Metrics' chief executive, confine the offer to Europe and Latin America. Which of the two describes availability for a buyer based elsewhere is left open.

How the measurement is supposed to work

Refinery89 set out a sequence in three stages. Before a campaign starts, insights are used to define the brand's standing with its audience. Once the campaign is live, questionnaires built on Brand Metrics' technology capture user sentiment in real time across four measures: awareness, consideration, preference and purchase intent. The output, according to Refinery89, indicates which stage of the customer journey a user occupies and what is likely to move that person to the next one, so that messaging, creative and targeting can be altered during the flight.

Mid-funnel, the phrase both companies lean on, is borrowed from the conversion funnel, in which consideration sits between first contact with a brand and the intent to act. Conversion reporting concentrates on the bottom of that model, and brand studies have traditionally reported after a campaign ends. The pitch here is the space between the two.

Brand Metrics' own company description labels the fourth measure differently. Every result, measured or predicted, is reported against awareness, consideration, preference and action intent, according to Brand Metrics. Refinery89's statement and its spokesperson both use purchase intent. The labels may describe the same survey question, but neither document reconciles them, and for a buyer comparing studies across vendors the wording of the question is not a trivial matter.

The comparison base is the most concrete figure in the material. Results are set against more than 60,000 campaigns spanning 122 industry categories, drawn from more than 35 countries, according to Brand Metrics. A lift figure can therefore be read against category norms rather than in isolation - a design PPC Land also recorded when Utiq brought its Deterministic BrandLift product to Germany in October 2025 with comparisons against industry standards and competitor campaigns built in.

The statistical design is not described

Survey-based brand lift conventionally compares an exposed group with a control group that did not see the advertising. PPC Land set out that mechanism in its account of how Google's Brand Lift Studies work: a randomised control group is isolated, and a one-question survey reaches both groups roughly a day after exposure. The control group does the same job it does in a holdout study, supplying the counterfactual against which any change is judged.

Neither company explains whether a control group is used on Refinery89 inventory or how it would be selected, how many completed responses are needed before a reading is shown, or how often results refresh when the companies refer to real time. Those are not academic points. A study needs enough answers in both groups for a difference to be statistically defensible, which is why vendors have historically set impression floors.

The pre-campaign baseline raises a related question. Brand Metrics says agencies and advertisers use its prediction technology to know what a campaign will achieve before it runs, and to optimise it while it is live. Refinery89's statement does not say whether the baseline it describes comes from surveys fielded before the flight, from that prediction product, or from category averages.

In-flight optimisation carries its own tension. A study measures the effect of a given set of creatives delivered to a given audience; changing either midway changes what is being measured. How results are reported across such changes goes unaddressed.

What the executives said

Gijs Witteveen, Head of Commercial Operations at Refinery89, framed the integration as the step that makes brandformance measurable. "Thanks to Brand Metrics' technology, we can tell a brand exactly where its audience stands on awareness, consideration, preference and purchase intent, and benchmark that against thousands of other campaigns," he said. "That turns brandformance from a strategic idea into something brands can actually track and act on in real time, removing the guesswork."

He also described Refinery89's framework as one that "pushes users from being aware of a brand to actually connecting with it, ultimately being the driver for lower funnel results." No conversion, sales or retention data from Refinery89 campaigns accompanies that claim.

Anders Lithner, chief executive of Brand Metrics, placed the deal in a wider change in buyer expectations. "Advertisers no longer want to choose between building their brand and driving performance, they want to do both, and they want proof it's working while the campaign is still live," he said. "By embedding our brand lift measurement technology, Refinery89 is giving advertisers and agencies across Europe and LATAM genuine mid-funnel visibility across their campaigns."

Figures that arrive without sources

Two statistics in the material deserve scrutiny. Refinery89 states that 7 out of 10 brands are prioritising effectiveness as advertisers struggle to prove tangible outcomes. The figure carries a hyperlink, but the text names no study, publisher, sample size, fieldwork date or definition of effectiveness.

Brand Metrics' company description states that 95% of the world's top 100 advertisers have measured campaigns using its technology, followed by the parenthetical "(Source: Ad Age)". That attribution can be read two ways: as the source of the 95% figure, or as the source of the top 100 list against which the 95% share was calculated. Nor does the description say whether those advertisers commissioned studies themselves or whether publishers licensing Brand Metrics ran the measurement on campaigns they were selling. The distinction matters for a vendor that names publishers and broadcasters as users of its measurement technology.

Both firms also rely on superlatives. Refinery89 calls itself the "#1 partner for publishers across Europe and LATAM"; Brand Metrics describes itself as the world's leading brand outcomes platform, combining the world's largest brand outcomes measurement capability with prediction technology. Neither supplies a ranking method, revenue figure or client count to support those positions. Neither mentions accreditation by the Media Rating Council or any other independent audit of the survey methodology.

Measurement sold alongside the media

Refinery89's business sits mainly on the supply side of programmatic advertising. For publishers, the company offers a "Single Tag" and a "Monetize APP SDK", which it describes as a gateway to maximise ad revenue at scale. It lists itself as a Google Certified Publishing Partner and an active IAB member, a Google status that requires meeting specific standards for ad optimisation and implementation. Brandformance is the side of the same company that faces brands and agencies.

That structure sets this deal apart from most brand lift integrations PPC Land has tracked in 2026, which have placed the survey inside buying tools. LoopMe put its study into The Trade Desk's Measurement Marketplace on September 2, 2026, allowing buyers to configure a study during campaign setup without a separate contract or login. illumin embedded Cint brand lift into its programmatic campaign setup in May 2026, and Basis followed with Cint on June 16, 2026. In each case the measurement sits with the buyer's platform.

Here, by contrast, the party selling the impressions is also the party packaging the evidence of their effect. Brand Metrics is a separate company, and nothing in the material suggests Refinery89 controls the survey output. Still, Brand Metrics' own account of its customer base points to the sell side: publishers and broadcasters use its measurement to prove the value of their inventory, while agencies and advertisers are described as users of its prediction technology. The Refinery89 arrangement hands the measured product to advertisers through the seller.

The documents also leave the reader-facing side undescribed. Questionnaires appear to users of the network's sites and apps, yet neither company explains how respondents are chosen, where surveys are placed on publisher pages, or how consent is collected - questions with legal weight in European markets.

A crowded month for brand lift

The Refinery89 announcement is the third brand lift development in less than a fortnight. Twelve days earlier, LoopMe set its study floor at one million impressions, against competitor minimums of three to four million in many markets, according to that company. On September 9, 2026, Upwave opened a closed beta of Predictive Brand Lift, which estimates lift through modelling trained on observed studies across more than 3,000 brands, with no survey fielded at all.

That last development bears directly on Brand Metrics' positioning. Its company description refers to combining measurement with prediction "under one outcomes standard". Refinery89's statement describes questionnaires. It does not say whether predicted figures will appear alongside measured ones in Refinery89 reporting, or how they would be labelled. PPC Land's Upwave coverage made the point that observed and predicted lift answer the same question with different evidentiary standing.

Thresholds are another axis of competition. Utiq priced Deterministic BrandLift from campaign budgets of 50,000 euros when it went live in Germany on October 8, 2025. Brand Metrics says its technology delivers consistent measurement "even on the smallest campaigns", but neither document attaches a number to that.

The platforms have been moving too. Google exposed 24 conversion lift metrics and a set of brand lift dimension resources in Google Ads API v25.1 in August 2026, though only for allowlisted accounts. Cint, meanwhile, merged brand lift and sales lift into a single in-flight dashboard in June 2026, using transaction data for the second signal - an attempt to close precisely the gap between attitude and purchase that Refinery89's framing asserts but does not measure.

Brandformance as a sales label

The term itself is not new. Teads applied it to its Connected Ads format in 2025, pairing awareness-building placements with conversion mechanisms on the same page. IAB Spain's 2026 digital roadmap, published in January, described brandformance strategies as a way to connect branding and performance rather than treating the two as competing budget lines. Refinery89 uses the word both as a product name and as what its boilerplate calls a framework.

What evidence supports the premise? Some exists. Research by Tracksuit and TikTok found high-awareness brands converting at 2.86 times the rate of low-awareness brands on TikTok. That model, though, rested on 11 brands and 60 data points after filtering from an initial 147 brands in Australia and New Zealand. The link between brand metrics and sales is widely asserted and modestly evidenced.

Precision about what a survey records matters here. Brand lift captures what respondents say about a brand. Incrementality testing measures behaviour caused by advertising. As PPC Land's LoopMe coverage put it, neither answers the other's question. A rise in stated purchase intent on Refinery89 inventory is a signal about attitudes; whether it translates into purchases is a separate empirical claim.

Buyer frustration gives both kinds of product an opening. IAB's State of Data 2026 report, published on February 7, 2026, found up to 75% of respondents rating attribution, incrementality tests and mix models as underperforming. IAB Europe's first Attitudes to Digital Advertising Report, from January 2026, found 70% of advertisers frustrated with performance measurement, with brand metrics rated important by 36% for evaluating connected television. Those are different measures from the unnamed 7 out of 10 figure in Refinery89's statement, and they are not interchangeable.

The Latin American angle

Refinery89's Latin American footprint places it in a market that has seen notable consolidation this year. MiQ bought the Latin American business of Adsmovil in March 2026, creating what MiQ described as the region's largest independent programmatic company, with more than 150 employees across 12 markets.

Platform brand lift has reached the region gradually. Amazon's Brand Lift product reached Mexico in September 2025, which made it the eighth country with access and the third in the Americas. LoopMe, for its part, argued in its September announcement that in many markets brand lift measurement is simply unavailable to advertisers. Refinery89 has not named the Latin American countries where studies are live, so the practical reach of its offer in the region cannot yet be assessed.

What the deal changes

For agencies buying open-web inventory in Europe and Latin America, the partnership adds another route to in-flight brand metrics, this one attached to the supply rather than the buying platform. Buyers already using LoopMe through The Trade Desk, or Cint through illumin or Basis, now face a question of consistency: a study run on Refinery89 inventory reports on Brand Metrics' four measures and its own comparison base, rather than on the question wording and norms of those other vendors.

For Refinery89, brand lift data offers a way to argue that its publisher network moves attitudes rather than merely delivering impressions. Because the data is produced on the seller's own inventory, its credibility with buyers rests on methodology that the September 14 announcement left undisclosed.

Lithner said Brand Metrics was "excited to see the partnership scale." The companies gave no timetable for that expansion.

Timeline

Summary

Who: Refinery89, an ad tech company working with publishers and advertisers across Europe and Latin America, and Brand Metrics, a brand lift measurement vendor. The named spokespeople are Gijs Witteveen, Head of Commercial Operations at Refinery89, and Anders Lithner, chief executive of Brand Metrics.

What: A partnership that places Brand Metrics' survey-based brand lift measurement inside Refinery89's Brandformance offer. Questionnaires record awareness, consideration, preference and purchase intent while campaigns run, with results compared against more than 60,000 campaigns across 122 industry categories and more than 35 countries, according to Brand Metrics. Pricing, minimum campaign size, market list and statistical design were not disclosed.

When: The partnership was disclosed on September 14, 2026. The companies said it had already rolled out, without giving a start date, and that it is now expanding.

Where: Across Refinery89's publisher network in Europe and Latin America. The announcement also refers to advertisers worldwide, a framing the rollout details do not support.

Why: Advertisers face pressure to show what brand spending achieves before a campaign ends, and survey-based lift has been moving into the platforms where media is bought. This deal places the measurement with the seller of the media instead, which gives Refinery89 a way to argue for its inventory while leaving buyers to judge methodology the companies have not yet published.