This trailer sells the cast, not the story. The audience is buying anyway.
Avengers: Doomsday and Dune: Part Three open the same day, December 18, 2026. We showed the public Doomsday trailer to two separate samples: 26 AI-moderated conversations and a 111-person survey. Five findings, two bonus deep-dives, and the receipts at the bottom.
26 CONVERSATIONS111 SURVEYS1 TRAILER
INSIGHT ONE OF FIVE
THE MOST LOPSIDED NUMBER IN THE STUDY
The returning X-Men cast sells the movie
90of 111 survey respondents (81.1%)
said the legacy X-Men cast, Patrick Stewart, Ian McKellen, Rebecca Romijn, increased their interest in the movie. One respondent said it decreased their interest. Nobody said it decreased significantly. Each square below is one person.
"the fact that he's in as a superhero is just [chuckles] amazing."P12, mid-engagement, 55–64, Sacramento CA [19:39], on Patrick Stewart as Professor X
"It would be a nine now since I saw the introduction of the Fantastic Four characters. And colliding the X-Men universe with the MCU universe... I feel like is a really good move."P03, mid-engagement, 35–44, Vancouver WA [14:15]
INSIGHT TWO OF FIVE
MEANWHILE, THE PLOT...
Hardly anyone followed the story
23of 111 (20.7%) understood the stakes "very clearly"
The largest group landed in the vague middle. In the 26 conversations, nearly everyone could restate the surface premise (heroes team up, world's at stake), but almost nobody could say what Doctor Doom actually wants, or why this threat is supposed to be worse than Thanos.
"Based on the trailer, how clearly did you understand what this film is about and what is at stake?" All 111 answered.
"I don't think he's Thanos level strong. So I feel like Thor might have misspoken a bit by saying that he was more scared of the current threat than anything prior because Thanos should be scarier."P01, core/committed, 25–34, Grandview WA [14:47]
"right now you just have a villain that's doing something, and we don't know why he's doing it."P10, core/committed, 45–54, Pittsburgh PA [13:38]
INSIGHT THREE OF FIVE
WHO ACTUALLY MOVED
The die-hards were already in. The trailer worked on the middle.
Pre- and post-trailer excitement on a 5-point scale, by MCU familiarity. The core fans started near the ceiling and barely moved. The "watch occasionally" tier is where the needle actually moved.
Paired means, same respondents before and after watching. "Rarely/never" (n=4) shown dashed: base too small to read as a trend. Hover a line for exact values.
"Yes and yes. I've already bought my ticket for the movie."P05, core/committed, 25–34, Danbury CT [17:31]
"Oh my goodness, the world's gonna end. Now the world's gonna end extra bad. Now the world's gonna end times twenty."P24, mid-engagement, 35–44, La Mesa CA [07:17]
INSIGHT FOUR OF FIVE
THE DATA DISAGREES
Does the villain feel dangerous? OPEN QUESTION
The study's two lenses give opposite answers. Flip between them.
Recognition without menace. The most comics-literate voices, independently and unprompted, said the trailer never demonstrates Doom's power:
"they did not make him feel dangerous or powerful, and that was definitely missing."P15, mid-engagement, 55–64, Haverhill MA [13:36]
"in previous instances of Doctor Doom being in films, he just wasn't that overpowered."P17, core/committed, 18–24, College Station TX [06:53]
"it kind of feels like they're fighting each other more than they're fighting him. That's what the trailer seemed to show me."P22, mid-engagement, 35–44, Carol Stream IL [14:31]
Presence-based read from a 26-person sample. Not sized for population-level inference.
Majority-positive. At N=111, the aggregate says the opposite:
61.3%
rate the villain "very" or "extremely" threatening
68 / 111
59.5%
say his presence increased their interest
66 / 111
37.8%
say his presence had no effect
42 / 111
And recognizing who he is at all is a near-even three-way split:
37.8%
35.1%
27.0%
knew exactly 42 / 111general idea 39 / 111no idea 30 / 111
The public trailer never reveals who is under the mask. (The two 42/111 counts on this panel come from different questions and happen to match.)
A strong reveal can create a positive halo that a quick survey scale picks up, while the interviews' follow-up questions cut through it for a small, highly engaged group. Two methods, two answers, both real.
INSIGHT FIVE OF FIVE
QUAL-ONLY, N=26 · NOT SIZED AT SCALE
One crossover, two very different welcomes
In the conversations, the returning X-Men legacy cast reads as nostalgic even when people are complaining. The new Fantastic Four cast draws real resistance, including from committed fans. This is a voices finding, not a percentage: the survey never asked a Fantastic-Four-specific question, so the asymmetry can't be sized at N=111.
X-MEN LEGACY
warm, even inside the complaints
"It would have felt better if they had have actually gotten Halle Berry and Famke Janssen to, to be in the film... Uh, Shawn Ashmore as well... they couldn't even get the costume right for Mystique."P10, core/committed, 45–54, Pittsburgh PA [18:18]. Even the complaint is a wish for more legacy cast, not less.
"X-Men? Um, I didn't even notice. I, I didn't notice. But now that I know, I mean, that's pretty cool, too, you know."P14, casual/occasional, 35–44, Atlanta GA [14:44]. Recognition arrives late, still lands positive.
FANTASTIC FOUR
cooler, sometimes outright resistant
"I just don't like Pedro Pascal as, well, Reed Richards. I think it was a horrible casting decision, and it just-- it actually just ruined it for me."P09, casual/occasional, 55–64, Los Angeles CA [07:43]
"I felt like the Fantastic Four were making the screen too cluttered. I did not want to see them."P16, core/committed, 18–24, Charlottesville VA [11:14]
The survey can confirm the X-Men-legacy half on its own (the 90 of 111 in Insight One). The Fantastic Four comparison is a coverage gap for the next wave.
BONUS · THE DUNE SHOWDOWN
FORCED CHOICE, ONE PICK ONLY
A blowout that hides a coin flip
Overall, 76 of 111 (68.5%) picked Doomsday over Dune: Part Three for opening weekend; 18 picked Dune, 17 picked neither. The 26-person interview sample agreed (21 of 26, 81%). But step through the Dune-familiarity tiers below and watch the lead collapse.
DOOMSDAY
VS
DUNE: PART THREE
DOOMSDAY BELOW 50%
68.5% chose Doomsday31.5% Dune or neither
"Never heard of Dune" (n=4) shows a raw count only, too small to size as a percentage, per the source data's own suppression rule.
BONUS · MEET THE 26
EVERY INTERVIEW, ONE DOT
Read them in their own words
All 26 interviews, arranged by MCU familiarity and quality-scored. Click any dot to read that participant's exact words, with the timestamp.
← Click any dot to see that participant's segment, age, geography, quality band, and their own words.
THE RECEIPTS · QUALITY, METHODS & WHO WE TALKED TO
HOW THE INTERVIEWS WERE SCORED
All 26, scored and shown
Every conversation carries an Interview Quality Score: Signal Depth (50%), Moderator Effectiveness (30%), Data Usability (20%), each 1 to 5. Study-level mean: 3.98 / 5. One tile per interview, colored by band; hover for detail.
Exceptional
Strong
Adequate
Thin
flagged
Three interviews (P10, P14, P22) had the trailer's own audio picked up on the participant's track for a stretch; that stretch was excluded from coding, the rest kept. Six participants disclosed they'd seen the trailer before the session, which weakens only the before/after excitement measure for those interviews.
AND THE SURVEY?
The 111-person survey has no per-interview quality score: it's self-administered and closed-ended, so there's no interviewer or transcript to rate. Its quality signal is procedural instead, 100% item completion plus two near-duplicate frequency questions that agree for 102 of 111 respondents (91.9%), a sign of attentive responding.
WHO WE TALKED TO
Participants appear only as P## codes, with segment, age bracket, and general geography. No real names.
Interviews, N=26
26 distinct US cities and regions, both coasts and interior markets. No geographic quota.
Survey, N=111
The survey collected no age, gender, or geography data. There's no demographic breakdown to show for this sample.
ONE MORE PANEL · WORK WITH US
HAVE A QUESTION?
Partner with us on our next study
This wave left things on the table: the Fantastic Four comparison was never sized at N=111, and nobody has measured what any of this does to opening weekend. If there's a question you want answered in the next round — about this trailer, this audience, or your own category — leave your email and we'll bring you in on the design.
YOU'RE IN
Got it. We'll be in touch when the next wave goes into design, and you'll get the opening-weekend numbers before they're public.
One email, used to talk to you about research. No resale, no third-party lists.