Films are frequently altered after they are shot, and the ending is the section most likely to move. The process behind that is more procedural than the phrase audience testing suggests.
What a test screening actually measures
A near-finished cut is shown to a recruited audience that broadly matches the intended market. Viewers fill in cards afterwards, and a smaller group stays for a moderated discussion.
The cards produce comparable figures across screenings, including how many people would recommend the film. Those numbers travel further inside a studio than any individual comment does.
The discussion adds the reasoning behind the scores. A room can explain that it disliked a character's final choice, which the numbers alone would only register as a drop in enthusiasm.
Why endings absorb most of the changes
An ending is usually the smallest section to alter. Reshooting a final confrontation may require a few days and a limited set, where restructuring a second act would touch everything downstream.
Endings also carry disproportionate weight in how a film is remembered. A viewer leaving unhappy will describe the whole picture in those terms, which shapes what they tell other people.
Because recommendation is the metric studios watch most closely, the last ten minutes attract attention out of proportion to their running time.
The feedback is filtered before anyone acts on it
Raw audience reaction is not applied directly. Research staff look for patterns that repeat across screenings, since a single unhappy room may reflect the recruitment rather than the film.
A confusion signal is treated differently from a dislike signal. Confusion often points at a missing piece of information that can be fixed with a line of dialogue or a reinstated shot.
Where the process breaks down
Test audiences respond to what they have just watched, not to what a film might become on a second viewing. Deliberately unresolved endings and slow-building tones tend to score poorly by design.
Directors who work in ambiguity often push back for that reason, and the strength of that pushback usually depends on their contractual right to a final cut.
Films that were altered against a filmmaker's judgment sometimes reappear years later in a restored version, which is why alternative cuts exist at all.
The practice has adapted to leaks
Phones have made secrecy harder, and details from unfinished cuts now circulate before a film is complete. Studios have responded by tightening entry procedures and by screening in smaller markets.
Some productions now test with unfinished effects deliberately obscured, so that any leaked description carries little weight. The result is a process that has grown quieter without becoming less influential.