Why the Best TV Ads Have a Big Cast and Fewer Speaking Roles

Jacob Trussell | 7 Min Read

Why the Best TV Ads Have a Big Cast and Fewer Speaking Roles

TL;DR

Quick Takeaways

What the research suggests about who appears in your ad and who gets a line.

  1. In QuickFrame and MNTN’s study, 15- and 30-second ads with six or more actors had the highest visit and conversion rates.

  2. Ads performed best when about half the on-screen talent spoke, suggesting that a bigger cast doesn’t need more dialogue.

  3. Actor count may reflect production budget, so adding faces alone isn’t a proven way to improve performance.

  4. Female voices earned stronger trust ratings in the cited video and radio studies, though neither measured sales.

  5. Test casting and speaking roles against your own results, and try a voiceover swap as an inexpensive comparison.

Research suggests specific combinations of ad creative elements produce specific outcomes and sentiment.

Is there such a thing as too many actors in a television commercial? Or too many speaking roles? We investigated the creative elements that contribute to a high-performing commercial in Video Vitals: What Makes the Perfect CTV Commercial?, a report from QuickFrame and MNTN that analyzed more than 1,800 CTV ads across ad length, tone of voice, emotion, number of scenes, and more. Two of the attributes it coded were the number of actors on screen and the number of them who spoke. The two didn’t point the same way.

The number of actors versus the number of speaking roles

Casting talent can take up a chunk of a brand’s CTV commercial budget, yet advertisers have traditionally had difficulty quantifying the impact of casting decisions. The Video Vitals findings provided some insight here, and revealed that brands should lean toward including more talent in their commercials, as it has a positive correlation with both visit and conversion rates: both 15- and 30-second commercials saw peak performance when six or more actors were present throughout the creative.

Effect of Number of Talent on Visit and Conversion Rates

15-second and 30-second commercials: Visit Rate and CVR by number of talent15-second and 30-second commercials: Visit Rate and CVR by number of talent plotted with the original categories, series, and values.Average15-second Commercials−61%57%−9%−105%−2%−79%25%−34%−22%−26%−41%−15%−29%28%0123456+Number of TalentVISIT RATECVRAverage30-second Commercials−4%−60%32%−13%−17%−45%13%−56%−39%−61%−62%28%−4%56%0123456+Number of TalentVISIT RATECVR

Scroll horizontally to see full chart.

Number of Talent data
Number of Talent15-second Visit Rate15-second CVR30-second Visit Rate30-second CVR
0−61%−34%−4%−56%
157%−22%−60%−39%
2−9%−26%32%−61%
3−105%−41%−13%−62%
4−2%−15%−17%28%
5−79%−29%−45%−4%
6+25%28%13%56%

Before reading too much into that, a note on what the data is. All 1,800-plus ads ran through MNTN’s platform, and attributes like actor count and speaking roles were tagged by QuickFrame’s Video Vitals AI tool, with performance measured against the average across the full set. Actor count is also hard to separate from production budget. An ad with six or more people in it usually has more scenes, more locations, and more money behind it than an ad with one, and some of what reads as a casting effect may be a production-value effect. The correlation is real in this sample. Whether adding faces to an otherwise identical spot would move the numbers is a question the study can’t answer.

The same data cuts the other way on dialogue. Brands should pay as much attention to the number of speaking roles as to the number of cast members, because the “more is more” approach does not apply here. Connected TV commercials performed best when only around 50% of the on-screen talent had lines in a scene.

Effect of Number of Speaking Roles on Conversion and Visit Rates

Visit Rate and CVRVisit Rate and CVR plotted with the original categories, series, and values.Average−18%23%15%15%−13%−51%5%89%−4%−42%0%25%50%75%100%Percentage of Talent SpeakingVISIT RATECVR

Scroll horizontally to see full chart.

Percentage of Talent Speaking data
Percentage of Talent SpeakingVisit RateCVR
0%−18%−51%
25%23%5%
50%15%89%
75%15%−4%
100%−13%−42%

One reasonable read is that CTV is watched from across a room, often with half an eye, and a commercial that splits its message across four voices is harder to follow than one that gives it to two. Crowd scenes show the product in use. Dialogue has to carry the point. Giving the second job to fewer people may simply keep the message clear. The study records the gap without explaining it, so treat that as one plausible reading, and a cheap one to test.

Who gets cast, and who gets to speak

Casting a certain number of talent can work in advertisers’ favor, and other research suggests the type of talent matters too. Take gender. A 2017 study from J. Walter Thompson and the Geena Davis Institute on Gender in Media analyzed more than 2,000 English-language ads from a decade of Cannes Lions entries and found twice as many male characters as female characters. Additionally:

  • 25% of ads featured only men, while 5% featured only women.
  • 18% of ads featured only male voices, while fewer than 3% featured only female voices.
  • Women in ads were mostly in their 20s, while men appeared across their 20s, 30s, and 40s.
  • Female characters were 48% more likely to be shown in a kitchen, and male characters were 50% more likely to be shown at a sporting event.

Two caveats. The sample is award-show entrants, which skews toward big budgets and big agencies, and the data runs through 2016. It’s a picture of the industry’s most celebrated work from that decade, and CTV creative today may look different. The pattern is still useful as a baseline. If the default commercial casts and voices men, a brand that does otherwise is at minimum standing out.

What the voiceover research shows

Although women have been underrepresented on screen, their voices don’t fall short where voiceovers are concerned. A study from CXL Institute, a marketing education provider, recruited 202 participants around the U.S. and showed them the same promotional video with four different voiceovers: a professional and an amateur, male and female. Participants rated each on trustworthiness and attractiveness. They trusted the female voices more, professionalism made no measurable difference to trust, and the professional male voiceover was rated least attractive.

Voiceover Trustworthiness and Message Attractiveness

Voiceover trustworthiness and message attractiveness. Unlabeled bar heights faithfully traced from the supplied chart; no exact scores inferred.Voiceover trustworthiness and message attractiveness. Unlabeled bar heights faithfully traced from the supplied chart; no exact scores inferred. plotted with the original categories, series, and values.How Trustworthy Would YouRate This Video?ExtremelyTrustworthyVeryTrustworthyTrustworthyNeutralUntrustworthyVeryUntrustworthyExtremelyUntrustworthyFemale(Pro)FemaleMale(Pro)MaleHow Attractive Is ThisVideo’s Message to You?ExtremelyAttractiveVeryAttractiveAttractiveNeutralUnattractiveVeryUnattractiveExtremelyUnattractiveFemale(Pro)FemaleMale(Pro)Male

Scroll horizontally to see full chart.

The pattern held when visuals were removed entirely. Research from Westwood One and Veritonic tested 164 award-winning radio ads with a panel of 6,003 consumers and found that ads with a female voice performed as well as or better than male-voiced ads on happiness, authenticity, trustworthiness, and uniqueness, and on par across every other emotional metric. Only 4% of the ads tested were primarily voiced by a woman; 70% were voiced by men.

Female Voiceovers Test As Well as Male Narrators

Average score by attribute by gender of voiceover talent in radio ad

Average attribute scores by gender of voiceover talent in radio adsAverage attribute scores by gender of voiceover talent in radio ads plotted with the original categories, series, and values.7071Happy6667Authentic6565Interesting6365Trustworthy6769UniqueMALEFEMALE

Scroll horizontally to see full chart.

Attribute data
AttributeMaleFemale
Happy7071
Authentic6667
Interesting6565
Trustworthy6365
Unique6769

Neither study measures sales. CXL tested a single video with a general-population sample, and the Westwood One work is radio creative rated on perception. What they share is a direction: across different formats and samples, listeners rate female voices as more trustworthy. For an advertiser, that’s enough to justify a test, and voiceover is one of the cheapest variables in a commercial to swap.

What to test on the next shoot

Three findings, each cheap enough to test before the next shoot. More actors on screen correlated with better visit and conversion rates, so the casting budget is worth spending. Ads performed best when about half the cast spoke, so the script doesn’t have to give everyone a line. And the research on voiceover gender is consistent enough to be worth a side-by-side. None of these is a rule. A single-presenter testimonial with one voice can beat a crowd scene for the right product. But casting is one of the few creative decisions where advertisers have mostly worked on instinct, and this is a place where the data can at least narrow the choices.

Stay Smart on All Things CTV

Get insights, must-know stats, and clever ad strategies straight to your inbox.