Research in Real Time: Findings from our 2026 World Cup Study

Two months later, now that football (or soccer) season has kicked off, we're reflecting on World Cup 2026. During the 2026 World Cup we ran 35 AI-moderated interviews on Strella with fans across six countries, fielded in under 60 hours with 23 completed on the first day. Because we talked to fans while the tournament was still running, we caught reactions that don't survive being asked about later: the emotion of a specific match still live, and a gap between what fans remembered and what they actually did.

The problem with asking people later

Most research about a moment happens after the moment. Not because the research wasn't important or the team didn't want to hear what people really think. It's the logistics, the cost, the time. Typically, the tournament ends, the campaign wraps, the ad goes live and a week or maybe 4 weeks later a study fields or a survey is sent out asking people to remember how they felt and what they did.

I can't begin to count the number of "what did you think about your experience with..." emails I receive each week!

By the time you get the study or survey back you are not measuring the experience. You are measuring the memory of it, which has been quietly edited. People compress a chaotic two weeks into a tidy story. They attribute a decision to a reason that sounds sensible. They forget the ad entirely, or they remember the brand and reconstruct a motive for it.

The specific thing that gets lost is the gap between what people felt, what they did, and what they say drove them. Those three come apart in real time and knit back together in hindsight.

When it came to the World Cup this summer, we wanted to see what stayed visible if we did not wait.

The World Cup Study Results

Thirty-five interviews with World Cup fans, fielded July 8 through 11, in the middle of the tournament. Participants were in Canada, the US, the UK, Mexico, Spain and France, and could speak in any of ten languages.

The whole field window was under 60 hours. Twenty-three interviews completed on the first day.

Strella handled the scheduling, the moderation (in every language), and the analysis on the back end. Interviews ran in parallel, at whatever hour suited the participant, in whatever language they preferred. Analysis was available as the interviews landed rather than weeks after fieldwork closed.

Run traditionally, a 35-person study across six countries and multiple languages is a four to six week project. We did this study in just a few days. Here are the findings:

Finding 1: see live emotion as it happens

The clearest sign we had caught people in the moment was how specific and physical their answers were.

Average self-rated emotional intensity about the most recent match watched was 7.7 out of 10, with 23 of 35 rating themselves an 8 or higher. Ten described a bodily reaction unprompted: heart palpitations, sweating, goosebumps, a held breath. Akinwale, a regular bettor in the UK, said simply, "Oh, there was so much palpitation of the heart, you know. I was sweating in the evening." Two described avoidance, one turning the television off, another walking away in the 86th minute.

About ten participants independently anchored on the same match, a comeback most of them had lost money on. They described where they were standing and who they were with.

Ask that question in September and you get "it was exciting." The texture is gone, and with it the ability to tell an intense reaction from a polite one.

Finding 2: reveal the “say” vs. “do” gap in real time

We asked two questions that sound like the same question. What advertising do you remember from the tournament? And did any of it change what you actually did?

Twenty-three of the 33 participants with usable data named a specific ad unprompted. Six said an ad had changed anything they did.

That is not a story about bad creative. Recall was working. People could name brands, describe spots, sing the music back. The ads were doing the job ads are usually measured on. They were not doing the job they were bought for.

Simon, a UK bettor who trades on exchanges, described the mechanism better than our analysis did: "No adverts has kind of made me do a promotion. I'll have already done it because I've already seen it." The ad was not a trigger. It was a reminder of a decision he had already made.

The reason this matters beyond one category is that most post-campaign measurement asks the first question and treats the answer as a proxy for the second. Recall is easy to collect and it feels like evidence. Here it pointed the wrong way, and the only reason we could see that is that both questions were asked in the same conversation, close enough to the behavior that the second one had real detail in it.

Finding 3: uncover what customers really want

Asked what would keep them on a betting platform, 19 of 35 named promotions, bonuses or offers. It was the loudest request in the study by a wide margin.

Nine had genuinely switched apps because of a promotion at some point, so offers do work. But almost every switch they described was months or years old. Only one described changing anything during the tournament itself.

So the same people who largely ignored promotional advertising, and largely did not move on an offer, named offers as the thing that would keep them.

This is not people being dishonest. It is the difference between the answer that is easy to produce and the thing that actually happened. Asked what they want, people name the most available concept. Asked to walk through what they actually did last Tuesday, they describe something else entirely.

Every category has its version of this. Ask shoppers what would make them switch supermarkets and they say price. Ask what they did last week and it is about parking, or a habit, or which one is on the way home.

What in-the-moment research gets you that a later study cannot

Reactions with detail still attached. Not "it was exciting" but where they were standing and what their hands were doing. Detail is what lets you tell a real signal from a polite one.

Behavior close enough to recall accurately. Every participant could describe what they actually did that week. Ask the same question in the autumn and you get a plausible reconstruction.

The format also matters in the moment. Dynamic follow ups from Strella allowed participants to open up and share, uncovering trends and findings that no one was looking for:

• Trust created new strategies: Six participants raised match-fixing or officiating bias unprompted, and one had turned that belief into a rule about which matches he will not bet on.
• Education as a barrier to betting: four participants described a knowledge barrier keeping them from betting more, including one who sat out the tournament entirely because he could not read the lines.

Neither would have come up in a traditional "post event" survey with fixed answer options. Both findings have implications how you might reach these customers.

We ran this study on a football tournament because we were genuinely energized by the World Cup. In the moment research isn't unique to the World Cup or football. Any moment with a spike attached has the same shape: a retail holiday, a streaming launch, a product release, a conference, a season premiere. The spend concentrates, the measurement happens afterward, and the gap between what people recall and what they did closes over quietly in the retelling.

The full readout

The complete readout covers the method and sample, the full section-by-section findings including match-day ritual and how people actually make a pick, and more of the verbatims.

Frequently asked questions

What is real-time or in-the-moment customer research?
Research fielded while an event or experience is still happening, rather than weeks afterward. The advantage is that participants describe what they actually did and felt rather than a reconstructed memory of it. It requires recruiting, moderating and analyzing within days, which is why it is usually run with AI moderation.

How fast can you run a multi-country qualitative study?
Our World Cup study completed 35 interviews across six countries and multiple languages in under 60 hours, with 23 of them finishing on the first day. Interviews run in parallel rather than being scheduled sequentially, which is what compresses the timeline. Run traditionally, the same study is typically a four to six week project.

Is ad recall a good measure of whether advertising worked?
Not on its own. In our study, 23 of 33 participants named a specific ad unprompted while only six said an ad changed anything they did. Recall measures whether a message landed, not whether it moved anyone. The two only separate if you ask both questions of the same person, close enough to the behavior that they can describe what they actually did.

Why does timing matter so much in qualitative research?
Because memory edits. People compress a long experience into a tidy story and attribute decisions to reasons that sound sensible in hindsight. In our study, the gap between what fans recalled, what they did, and what they said would keep them was only visible because all three were asked while the tournament was still running.

Strella helps teams talk to their customers while the moment is still live. AI-moderated interviews in 40+ languages, with analysis in days instead of weeks. See Strella in action.