The Act Of Choosing What To Watch Boosts Recall of Content

As TV viewing was moving away from the "river" model of traditional television to the à la carte model of streaming and downloading, we began to study the effects this new model might be having on traditional advertising metrics.

In one experiment, we let a group of people choose a trailer they would watch, and compared it to a group who didn't have such choice before watching it.

Here's how we set it up.

The experiment was conducted online. Each group consisted of about 120 people who indicated they had not seen the trailer or the movie prior to the experiment.

On the first screen, we set up the experiment. Respondents were randomly assigned to one of the two groups. Respondents in the experimental choice condition (Group 1) were presented with two different descriptions for a movie trailer they hadn't seen, and asked to choose one. 

Participants assigned to Group 2, the control no-choice condition, were simply informed they were about to watch a trailer.

On the next screen, participants in the first group received an acknowledgement of their "choice", and both groups were informed of what they were about to watch.

Then, each group watched the trailer -- the same one!

Finally, all participants answered a set of experimental questions disguised among noise questions. Because of trailer's nature, we were able to collect the kinds of metrics that usually interest advertisers and marketers, such recall of the brand name or of the phone number that appeared on the screen.

The Results

For all six questions, more respondents in the choice group answered correctly than in the control group.

The significant differences were:

• Correct recall of the main character's name: 60.9% vs 47.7%
• Correct recall of the phone number on the screen: 52.6% vs 39.4%
• Correct recall of the car's brand name: 27.1% vs 17.4%

And a finding we had not anticipated:

Participants in the choice condition overwhelmingly (82%) chose to watch the trailer for "a comedy about a man looking for a reversal of his fortune" rather than "a comedy about a disgraced businessman who lost his family."

Confounding variables:  we have not investigated the possibility that the difference between the groups could be explained by the exposure to the trailer descriptions in Group 1 vs no exposure in Group 2. That is, priming Group 1 participants with the information about the trailer could have led to an increased recall of facts.