Marmite's latest out-of-home campaign, titled "Love it. Hate it. Cook with it.", generated eight times more disgust than the average UK outdoor advertisement when measured by creative effectiveness platform System1 on 27 April 2026 - a result that, paradoxically, may be exactly what the brand intended. The data, published on 5 May 2026 according to System1, places the creative in unusual territory: high disgust, high happiness, low short-term sales signal, and a Star Rating that sits squarely in the middle of the UK outdoor category.

The campaign is one of the most technically revealing pieces of outdoor creative to be put through System1's Test Your Ad Outdoor platform in recent months. The numbers tell a layered story - about how polarising brands manage emotional range, about the limits of standard creative effectiveness benchmarks, and about what happens when a brand deliberately courts negative feeling.

What the Test Your Ad data actually shows

The System1 report, dated 27 April 2026, tested the Marmite campaign against a nationally representative UK sample of 150 respondents. The overall Star Rating came back at 2.0, which places the ad in the "Modest" band. To understand what that means in context: according to System1, 30.1% of UK outdoor image ads score "Low" (below 2.0 Stars), 40.1% score "Modest" (the band this ad sits in), 21.4% score "Good", 6.2% score "Strong", and 2.2% score "Exceptional". The category average for UK outdoor image ads is 2.5 Stars. Marmite's creative comes in below that average.

The Spike Rating - which measures short-term sales potential by combining brand recognition strength with emotional intensity - registered 0.82, landing in the "Low" band. The category average for this metric is 0.96. A score below 1.0 on the Spike Rating suggests the ad is unlikely to drive an immediate uplift in purchase behaviour on its own. That is a meaningful data point for a brand whose outdoor placements are presumably competing for attention against food and drink competitors aiming for exactly that kind of direct activation.

The Fluency Rating, which measures how strongly the ad establishes brand recognition, came back at 76 - classified as "Modest." Within that metric, the platform also reports a Fast Fluency score, which measures how many people recognised the brand within two seconds. Here the Marmite ad scored 37, against a category average of 58. In other words, a significantly smaller proportion of viewers identified the brand quickly on first contact than the typical UK outdoor image ad would achieve. That is a notable gap.

The emotional breakdown

The emotional composition of the response is where the data becomes genuinely interesting. According to System1's report, 27% of respondents felt happiness in response to the campaign. That is broadly in line with the UK outdoor image category average of 30%. The disgust figure, however, is far from average: 8% of respondents reported disgust, against a category average of 1%. That eightfold difference is the headline statistic System1 highlighted when discussing the campaign. A further 1% registered fear - an emotion that effectively never appears in standard outdoor food advertising.

The report also breaks out contempt and anger readings. Contempt came in at 6%, against a category average of 5%, while anger registered at 0% - level with the category. Neutral responses accounted for 46%, versus the 52% category average. The data suggests the campaign pushed a meaningful portion of its audience toward stronger emotional territory in both directions. Fewer people felt nothing.

The happiness breakdown - which System1 analyses using its types-of-happiness taxonomy - is worth examining in detail. The system ranks happiness types by their ability to drive business effects. Top-performing happiness types include being "Amused" and feeling "Sensory Pleasure." The data from this campaign shows notable indexing on schadenfreude and amusement, which aligns with the brand's long-standing comedic identity built around the love-hate divide.

Among the verbatim responses collected by System1's MindReader tool, 23 respondents agreed with the sentiment "I love Marmite" as a reason for happiness. A separate group of nine respondents cited the sentiment that the product "gives that little bit of extra oomph to my dishes" - a response that points to the campaign's secondary objective: repositioning Marmite as a cooking ingredient rather than simply a breakfast spread.

What the key associations reveal

System1's report includes a ranked list of key associations - the first things respondents connected with the ad. Pizza came first at 51%, followed by Marmite itself at 44%. "Tasty" appeared third at 26%, then "Food" at 25%, "Spread" at 21%, "Lovely" at 15%, "Disgusting" at 15%, and "Nasty taste" at 13%. The co-presence of "Lovely" and "Disgusting" among the top eight associations, each at 15%, illustrates the campaign's unusual emotional symmetry. Most food advertisers would consider "Disgusting" appearing in 15% of free-association responses to be a serious problem. For Marmite, it is arguably the point.

The distinctive brand assets data adds another dimension. According to the report, 30% of respondents identified the brand through logo or symbol. Colour and font each registered at 23%, product or packaging shape at 23%, and slogan or tagline at 15%. Notably, 37% of respondents said they could identify the brand through "None of these" named assets - suggesting a portion of recognition came from diffuse brand familiarity rather than any single identifiable element. Only 5% identified a character or mascot.

Demographic differences in response

The System1 data breaks the audience into gender and age cohorts, and the divergence is substantial. Men gave the campaign a Star Rating of 2.4, which edges into the territory between Modest and Good. Women gave it 1.6, placing it in the "Low" band. That is a meaningful gender gap and would typically prompt creative review.

The age breakdown follows a different pattern. Respondents aged 18 to 34 gave a Star Rating of 2.3, though System1 flags the base size of 41 respondents in this group as below their threshold for statistical confidence. The 35 to 54 cohort gave a rating of 2.6 - the strongest of any group. Those aged 55 and over gave just 1.3, the lowest reading in the sample. That inverted pattern - strongest performance with mid-age adults, weakest with older consumers - may reflect how the "Cook with it" framing and the visual repertoire of pizza and pasta lands differently across life stages.

Fluency scores show an opposite pattern. The 55-plus cohort gave a Fluency Rating of 91, classified as "Strong," suggesting older viewers are highly effective at identifying the brand even if they respond to the creative less positively. Brand familiarity and emotional resonance appear to be pulling in different directions for that age group.

The strategic logic behind polarising creative

"This campaign is built on a brilliantly simple insight. Most people still pigeonhole Marmite as a breakfast staple, something for toast or crumpets, but here the brand challenges that by showing up in unexpected, everyday meals and reframing it as a versatile ingredient. You might not like it on bread, but you might love it in a stir fry. That tension is where the magic happens. At System1 we see that negative emotion, as well as positive, can drive short-term action, and Marmite strikes that balance well. There is always a risk of pushing too far into disgust, but for a brand that already divides opinion, leaning into that controversy is a smart way to cut through and drive trial," said Andrew Tindall, Chief Growth Officer, Advertising at System1.

The point about negative emotion driving short-term action is technically significant. Standard creative effectiveness frameworks are built around maximising positive emotional response - the assumption being that happiness correlates with purchase intent. System1's own research, published in its report "The Cure for Dull," argues instead that emotion of any kind - including negative emotion - outperforms neutrality, and that the most commercially damaging creative is advertising that generates no emotional reaction whatsoever. Under that framework, a Spike Rating of 0.82 for Marmite may be less damning than it initially appears, because the low score is partly produced by the high rate of negative emotion compressing the overall intensity metric.

This framing has broader implications for food and beverage advertising in the UK. Since 5 January 2026, England has operated under comprehensive restrictions on marketing foods high in fat, sugar, or salt - a regulatory shift that PPC Land covered in detail. Marmite is a yeast extract spread rather than a confectionery or snack product and sits in a different regulatory category. But the broader context of food brands navigating tighter rules on product-focused advertising makes the "Cook with it" campaign's brand-focused creative strategy - showing Marmite as an ingredient embedded in familiar dishes rather than as a standalone product - worth noting.

How this fits into System1's broader methodology

PPC Land has tracked System1's Test Your Ad platform across several recent campaigns. On 9 April 2026, System1 added AI-assisted summaries and a predictive measurement tool to Test Your Ad, introducing features designed to give creative teams faster reads on emotional response data. The Marmite report was produced using the existing FaceTrace methodology, applied to over 10 million people worldwide according to System1.

For context on category norms, PPC Land has reported on other UK outdoor campaigns tested through System1. Burger King's "Finished" OOH creative, tested on 22 April 2026, scored 4.3 Stars in its custom audience sample and 2.4 Stars with a nationally representative group - still higher than Marmite's 2.0 NatRep rating but illustrating how the same platform generates a wide range of scores across different creative executions. Lindt's Easter 2026 outdoor creative, tested on 30 March 2026, scored 5.9 Stars - the high end of the scale. The Marmite result, at 2.0, sits toward the lower end of the food and beverage outdoor creative distribution within this dataset.

System1 maintains a database of over 100,000 ads and operates in 81 markets globally, with offices in the UK, Europe, North America, Brazil, Singapore, and Australia. The Test Your Ad Outdoor variant applies the same FaceTrace emotional measurement framework used for TV and digital creative, adapted to the viewing conditions and dwell times typical of outdoor placements - in this case a simulated seven-second exposure.

What the OOH market context adds

Out-of-home advertising in the UK and globally has been expanding steadily. US OOH ad spend is projected to reach $4.0 billion in 2026, rising 4.1% year-over-year according to Guideline data reported by PPC Land in March 2026, with digital OOH formats growing at 14.5% while traditional formats grow at 1.5%. Research published in October 2025 by Keen Decision Systems, covered by PPC Land, found that OOH advertising delivers $7.58 in marginal ROI - outperforming many saturated digital channels - with food and beverage specifically reaching $7.72 per dollar spent.

Within that landscape, the question of how to measure creative effectiveness for outdoor formats remains contested. A campaign with a 2.0 Star Rating but an 8x disgust multiple sits awkwardly in standard performance frameworks. Does the disgust response drive recall and distinctiveness in a way that the Star Rating fails to capture? Does the happiness response - 27% of viewers, concentrated in specific demographics - indicate an engaged core audience that translates into commercial behaviour? System1's own methodology offers a partial answer: the platform scores ads on five dimensions covering emotion, fluency, visuals, communication, and attention. The Marmite creative scored in the Modest band across the first three and generated a dwell-time brand recognition rate of 76% - meaning three in four viewers identified the brand within a seven-second simulated exposure.

Why this matters for marketers

The Marmite campaign data matters beyond a single brand's outdoor execution. It surfaces a genuine methodological tension in creative testing: standard effectiveness benchmarks are calibrated against the average, and a brand that has spent decades building its identity on being above-average in polarisation cannot be assessed straightforwardly against those norms. Marmite's Spike Rating of 0.82 is below average. Its disgust rate is 8x above average. Both figures are accurate. Which one is more predictive of commercial outcome for this specific brand and category is a harder question, and one System1's platform - designed for cross-category comparison - is not built to resolve on its own.

For marketing practitioners, the data raises a practical question about how to commission and evaluate OOH creative for heritage brands with strongly divided audiences. The campaign's association data, where "Pizza" registers at 51% and "Marmite" at 44%, suggests the brand placement within the imagery is working at the recognition level even if FaceTrace positivity scores land below average. The juxtaposition of familiar comfort food with a divisive ingredient - the visual syntax of the campaign - appears to be communicating the "Cook with it" extension clearly. Whether that communication converts to trial among the non-users Marmite is presumably targeting is a question the outdoor test cannot answer.

Timeline

  • 27 April 2026 - System1 tests the Marmite "Love it. Hate it. Cook with it." outdoor campaign creative with a nationally representative UK audience of 150 respondents, generating a Star Rating of 2.0, Spike Rating of 0.82, and Fluency Rating of 76
  • 5 May 2026 - System1 publishes findings from the Test Your Ad Outdoor assessment of Marmite's campaign, reporting 8% disgust (8x the UK outdoor image category average of 1%) and 27% happiness among respondents
  • 5 May 2026 - Andrew Tindall, Chief Growth Officer, Advertising at System1, comments on the campaign's use of negative emotion as a creative strategy for a polarising brand

Related coverage on PPC Land:

Summary

Who: Marmite, the UK yeast extract brand owned by Unilever, whose "Love it. Hate it. Cook with it." outdoor campaign was assessed by creative effectiveness platform System1 using a nationally representative UK sample of 150 respondents. Andrew Tindall, Chief Growth Officer, Advertising at System1, provided commentary on the findings.

What: System1's Test Your Ad Outdoor platform generated a set of creative effectiveness metrics for the Marmite campaign: a Star Rating of 2.0 (Modest, below the 2.5 category average), a Spike Rating of 0.82 (Low, below the 0.96 average), a Fluency Rating of 76 (Modest), and an emotional response profile in which 8% of respondents reported disgust - eight times the 1% category average - while 27% reported happiness. The campaign ran imagery showing familiar dishes such as pizza and pasta alongside Marmite jars, aiming to extend the brand's positioning beyond toast and breakfast applications.

When: The System1 test was conducted on 27 April 2026. The data was published on 5 May 2026.

Where: The campaign ran as UK out-of-home advertising. System1 tested it under its UK Outdoor Image benchmark norms, using a seven-second simulated exposure via the Test Your Ad Outdoor platform. System1 operates in 81 markets globally, with UK operations contributing to its database of over 100,000 tested ads.

Why: The campaign's primary strategic objective, according to System1's analysis, is to reposition Marmite as a cooking ingredient used in a wider range of everyday meals - not just as a spread for bread. The use of an emotionally polarising creative strategy is consistent with Marmite's long-standing brand identity, which treats divided consumer opinion as a feature rather than a problem. System1's own research indicates that negative emotion, as well as positive emotion, can drive short-term commercial action, making the high disgust reading a calculated rather than accidental outcome within this particular brand's creative framework.

Share this article
The link has been copied!