The Hollywood Reporter dropped a bombshell behind-the-scenes exclusive on Supergirl, and it confirms what we’ve been reporting all week: DC knew this movie wasn’t working long before it hit theaters, and the fingerprints on the final cut belong to James Gunn.
UPDATE: A closer read of that THR post-mortem suggests someone at Warner’s may have leaked the story to pin the failure on Gunn, three weeks before Warner Bros. changes owners.
Per THR’s Borys Kit, the studio and director Craig Gillespie knew the movie wasn’t working as early as fall 2025, months after filming wrapped. Test scores never escaped the 60s across at least four screenings, with a top score of 70.
For context on how bad that is, THR notes Supergirl tested in the same range as Shazam! Fury of the Gods and Batgirl. One of those was a box office disappointment. The other one Warner Bros. decided was better burned for a tax write-off than released.
The studio took the movie over
After a December test screening came back just okay, the studio took charge of post-production and built its own cut. The moves that followed all point one direction.
DC brought in writer Jeremy Slater, a Gunn favorite, to help write scenes for a nine-day reshoot. If that name sounds familiar, Slater also wrote Mortal Kombat II, Warner Bros.’ other 2026 bomb, which collapsed to $128 million worldwide and got rushed to VOD inside a month. The writer brought in to help save Supergirl had his own WB tentpole play out the exact same way this summer, with the fans deciding to stay home.
The studio also added editor Fred Raskin, Gunn’s longtime cutter on the Guardians of the Galaxy movies and Peacemaker, alongside Gillespie’s editor Tatiana S. Riegel. THR describes Raskin as a key instrument of the studio’s slowly heavying hand.
And the big needle drop in the February cut, a cover of Cyndi Lauper’s “Girls Just Want to Have Fun” in the climactic set piece, was Gunn’s song choice before it was swapped for the Jimmy Eat World cover audiences are now dunking on.
Sources for THR describe creative differences between Gunn and Gillespie throughout, with one insider telling THR that “‘they were not creatively aligned’ is the polite way” of putting it. Other insiders push back, calling it normal filmmaker-studio friction.
And when THR politely says “the studio” took the movie over, read it plainly. THR’s own reporting defines the studio as the one run by Gunn and Safran, and the cut it built used Gunn’s writer, Gunn’s editor, and Gunn’s song. The studio cut was the Gunn cut.
The bakeoff where everyone lost
Here’s the stunning part. By deep winter, the scores had crawled into the low 70s. Instead of continuing to refine that cut, the studio forced a bakeoff in March, testing its own cut head-to-head against Gillespie’s version.
The scores dropped. Significantly. Both cuts.
The studio’s version edged out Gillespie’s by all of two points, and that’s the cut that went to theaters. Gillespie’s version, eleven minutes longer with more of the villain Krem, scored better on song choices, pacing, and the villain.
Read that again next to the opening-weekend complaints. Per Deadline, fanboys at the Burbank press screening called Krem two-dimensional. The needle drops got roasted. The cut that tested stronger on exactly those elements is the one that stayed on the shelf. The movie was never tested again, and from that point Gillespie had to fight for anything he wanted in his own film.
A franchise filmmaker not involved with Supergirl told THR that bakeoffs happen more than people think, but they aren’t normal, and a studio spending money on competing test cuts means it feels strongly about certain things.
It confirms what our insiders said
This lands almost exactly where our breakdown of why Supergirl failed already did.
Our Hollywood insiders told us the production churned through three composers because Gunn and Gillespie insisted on needle drops that couldn’t express the characters or the moments.
THR now confirms music was a key friction point between the two. Our insiders called the film underdeveloped and poorly written.
THR’s test scores, stuck in the 60s for months, say audiences agreed at every single screening. And the internal alarm we described, with the studio bracing while public tracking stayed inflated, now has a paper trail running from fall 2025 straight through a failed March bakeoff.
The result: a $37.1 million opening, below even Joker: Folie a Deux‘s $37.6 million, the previous benchmark for a DC bomb.
Gunn’s problem now
THR frames the aftermath as a scrutinizing spotlight on Gunn specifically, and notes there’s almost no modern precedent for a filmmaker directing his own movies while co-running the studio that has to fix everyone else’s.
Supergirl was the first DCU film Gunn didn’t write and direct, the test of whether the studio works beyond him. The studio’s answer to it failing in testing was more Gunn: his writer, his editor, his needle drop, his cut.
Per THR, Gunn and Safran’s contracts run through either the end of 2026 or the end of 2027 (full report here), with Clayface and Man of Tomorrow due before then, all of it under the shadow of the Paramount takeover we’ve been reporting on.
And buried at the bottom of THR‘s piece is a quote from a studio head that sums up the whole summer, that Gen Z doesn’t care about superhero movies, a genre that belongs to millennials.
Sound familiar? It’s the same generational shift we’ve been documenting while YouTube-born originals eat tentpoles alive at the box office.
DC didn’t just release a movie that failed. It released a movie it watched fail, in testing, for eight straight months.
