Kantar research commissioned by The Drum suggested that most of those who saw the Sydney Sweeney American Eagle campaign are likely to buy from the brand. Andrew Tindall says ‘Duh,’ claiming the results say more about memory and methodology than the work.
Congratulations if you’ve survived the Sydney Sweeney/American Eagle debate without screaming into your laptop. You’ve shown more restraint than the marketing internet, which has now climaxed with The Drum publishing data from Kantar that supposedly shows what Americans really think of American Eagle’s Sydney Sweeney ad.
Now, I love The Drum. It lets me write columns like this, so this is me biting the hand that feeds me. The article shared the survey results of 1,000 Americans who said they could recall the ad. Respondents were asked if the campaign made them more or less likely to consider or purchase from American Eagle in the future.
[Editor’s note: The research commissioned from Kantar and analyzed by The Drum provided an insight into how Americans were reacting to the news and social media cycle that developed around the American Eagle ad. It was not an evaluation of the ad’s creative effectiveness or performance for which Kantar has dedicated tools like LINK+ and LIFT+.]
The topline takeaway?
Most people who remembered the ad said they were likely to buy from the brand, suggesting the backlash won’t hurt sales. In said piece, foot traffic data from Pass By seems to disagree with the idea that they are buying.
This should reveal the reason I’m bothering to write this.
I genuinely care about using good data to find insights that transform marketing; it’s in my brand DNA. [Editor’s note: don’t mention genetics, FFS.]
This is not a beef with Kantar or a cheap shot; you can respond to me on these pages. I see this issue all the time, and we are cautious to avoid it. I don’t like seeing marketers convinced by shaky insights. It makes me uncomfortable.
So let me walk you through the mistakes. If marketers can’t spot when an “insight” is built on flawed methodology, we’ve got bigger problems than jeans ads.
The published data only detailed the opinions of people who could recall the ad.
That’s the classic Rosser Reeves Fallacy in play: if you only talk to people who already remember the campaign, you’re talking to a pool heavily skewed towards existing brand buyers.
This is such good marketing knowledge. Mainly because if you proclaim The Rosser Reeves Fallacy in a meeting, you sound like you know what you’re talking about. I try to explain this concept in a speedy 15-minute video here. [Editor’s note: 15?!?!]
Of course, heavy brand buyers will give you inflated positive sentiment: they were already predisposed to notice, recall and like the brand. It’s like only asking people leaving a Marvel premiere if they enjoy superhero movies. You’ve rigged the sample before you even hit “fieldwork.”
Another tip for corrupt brand managers: If you want to get your ad through approvals, select a sample that skews towards buyers. Things start looking rosy. Tip for brand managers worth their salt: If you spot someone trying to pull this one on you, call them out.
Measuring “future purchase intent” can be a slippery eel.
This question has been debunked repeatedly by decades of academic research. Humans are notoriously bad at predicting their own behavior.
We might say we’ll choose a brand next time, but in reality? Life intervenes. Competitors promote. We default to habit. As the behavioral science joke goes, if survey data were accurate, gyms would be standing-room only every January.
“Would you consider these jeans?” is not a predictor of what anyone will actually do.
It’s a vibes check, not an outcome measure.
And vibes don’t pay the rent.
It points to what we actually know about marketing. Decisions are made via subconscious memories and associations built over the years. As the great author of ‘Thinking Fast and Slow’, Daniel Kahneman, says, thinking to humans is like swimming to cats. They’ll do it if forced to.
Orlando Wood has researched this before, showing that emotion (Star Rating in the below) is a far better predictor of behavior over purchase intent (persuasion) in the IPA data.
But don’t get smug: marketers aren’t much better
Before we pile on, let’s admit our own track record isn’t great. Byron Sharp’s research has shown that even marketers, when asked to pick which ads will work, only choose correctly about 50% of the time.
Advertisement
That’s basically coin-toss territory. So, I agree with anyone who says we need to get out of the marketing echo chamber, which is why The Drum is bringing data to the debate.
We’re terrible proxies for real people. We live in an echo chamber, we obsess over viral campaigns most consumers never see, and we argue about semiotics on LinkedIn. It’s not exactly representative of the average Saturday morning denim shopper.
The bigger danger: sliding back to bad measurement
Here’s what really worries me, though.
Peter Field’s analysis of the IPA Databank has shown a massive shift in the quality of creative measurement over the past twenty years. It’s become better at predicting and explaining business results because we’ve moved away from these rational research methods to more emotion-first ones.
Two decades ago, creative pre-testing didn’t correlate with business effects. Now it does.
Why?
Because research has evolved.
We moved away from slow, rational, “System 2” metrics like purchase intent, persuasion and “message takeout,” and toward fast, emotional, “System 1” measures that track what people actually feel in the moment.
That shift is something marketers should be proud of. Creative testing works now because it focuses on emotion and memory (what actually drives brand growth) instead of asking people to rationalize answers they don’t even consciously know.
The rule of thumb: millions feeling > thousands explaining
So, here’s the takeaway:
Ignore outrage in the LinkedIn echo chamber (our opinions don’t matter).
Ignore surveys stacked with recallers (they inflate loyalty).
Ignore purchase intent questions (they’re behavioral fiction).
Instead, focus on what scales: how millions of consumers feel, not what a tiny sample can explain in words.
That’s what modern creative testing has cracked, and why it’s gone from “a nice-to-have” to an actual predictor of business effects. Ads that make people feel more go on to deliver more: market share, pricing power, profit.
Note from The Drum: If you are a methodology-arguing type, do join the debate. We only ask that you treat all the participants with respect. The Drum stands by the reporting and the research, flaws or not, the consumer gets the final say. We will be keeping an eye on those sales, the stock price and any subsequent research that follows.
Andrew Tindall is a senior leader in the advertising industry and SVP of Global Partnerships at System1, where he champions creative effectiveness and evidence-based marketing. Known for his sharp thinking on out-of-home, emotional advertising, and brand growth, he’s become a go-to voice on how creativity and data can work together. With experience across major brands like Diageo and Bacardi, Andrew now helps global agencies and marketers make better ads that actually work, contributing regularly to panels, podcasts, and the occasional punchy LinkedIn rant.
