Skip to content

Latest US poll: 63% say gas prices are affecting their household budget a great deal or a moderate amount

See the poll

Video ad testing

Test your video ad before the media spend

Show your ad to the US adults you want to reach and see whether they remember the brand, get the message and want to buy. Compare cuts on the same questions and hear the reasons in people's own words.

Talk to salesWatch a test run
3AI follow-up questions on each video answer, at most
5audience groups compared side by side, at most
309ready-made question templates, ad questions among them

Why test before media

Media is the expensive part

Most of an ad budget goes on airtime and placements, not on making the film. If viewers forget whose ad it was, or take away the wrong message, every placement carries that weakness. Testing the creative first shows which cut to run and what to fix while the edit is still open.

What can go on the bench

A finished filmA rough cutAn animatic30-second cuts15-second cuts6-second cutsVertical edits

Cut by cut

Every cut on the same questions

Each row is one measure, named the way the report names it.

Base: Three matched groups of US adults aged 18-44 who buy coffee away from home; each watched one cut.

Shares of viewers. Diagnostics shows "it was easy to follow". Feeling is the share who picked a happy or excited face.

Video ad testing · scorecardBrand A · video ad test report
Measure
30s30-second
15s15-second
6s6-second
Brand recallDid viewers name the right brand after watching? We also count how many named a competitor instead.
54%C9% named a rival
62%AC6% named a rival
47%12% named a rival
MessageWhat people say the ad was telling them, first in their own words, then from a list.
49%C
58%AC
33%
AppealHow much they liked the ad, on a five-point scale with every point labeled.
66%C
61%C
52%
PersuasionWhether the ad makes them more likely to buy the brand.
41%C
44%C
30%
FeelingHow the ad made them feel, picked from a row of faces. People choose for themselves; we do not read their faces.
52%C
48%C
37%
DiagnosticsAgreement with statements such as "it caught my attention", "it was easy to follow", "it fits the brand" and "it was irritating".
71%Ceasy to follow
78%Ceasy to follow
58%easy to follow
Likes and dislikesOpen answers on what worked and what did not, grouped into themes.
"Lovely, but long." "Lost the thread."
"Clear and quick." "Nice leaf."
"Too fast to catch."

Played to the end

Watch three cuts meet their viewers

  1. 01The ad plays to the end
  2. 02Which brand, which message?
  3. 03Scores build up, cut by cut
  4. 04The cut to run

Step 1 / 4

The ad plays to the end

Each viewer sees the whole ad before any questions. The Next button only appears when it ends.

Study set-up

Video ad testing

  • StimulusA 30-second film for the fall menu, and 15- and 6-second edits
  • WhoUS adults aged 18-44 who buy coffee away from home
  • DesignThree matched groups; each viewer sees one cut
  • ViewingThe whole ad first; Next appears when it ends
  • ThenBrand, message, feeling, persuasion, and the reasons

Set up with our team, or yourself in CatalyseAI

Step 2 / 4

Which brand, which message?

They say which brand it was for, what it was telling them, how it made them feel and whether they are more likely to buy.

Question 1 of 12

Which brand was this ad for?

The coffee place... Brand A?

How did the ad make you feel?

AnnoyedBoredNeutralHappyExcited
Next

"I lost it in the middle, then the cup came back."

"Short and clear. I knew it was Brand A."

"Six seconds was too quick to catch the name."

0:38

On-camera answer

AI follow-up: What was the ad trying to tell you?

"That the seasonal drinks are back. The leaf at the end was nice."

  • Live person
  • Light
  • Length
  • Relevance 0.86
  • Attention 0.91

Transcribed, checked and searchable. Up to three follow-ups per answer.

Step 3 / 4

Scores build up, cut by cut

Brand recall, message, appeal and persuasion for every cut, with markers where a gap is real.

Live tally, as answers arrive

30-second15-second6-second

Brand recall

  • 54%
  • 62%
  • 47%

Message

  • 49%
  • 58%
  • 33%

Persuasion

  • 41%
  • 44%
  • 30%

Step 4 / 4

The cut to run

The report sets the cuts side by side, so you pick a cut and fix it while the edit is still open.

Verdict

The 15-second cut runs

After a reworked end card, before the media is booked.

  • 15-second cutRun it
  • 30-second cutLiked, less remembered
  • 6-second cutBrand gets lost

The process

How a video ad test runs

From the cuts to the one that runs.

  1. 01

    Share the cuts

    Upload the finished film, a rough cut or an animatic. Tell us the brand, the competitors to list beside it and the message the ad is meant to carry.

  2. 02

    Pick the viewers

    Choose who should see the ad, for example US adults aged 18-44 who buy the category. We recruit them through PocketsFull, our own panel, and check the audience can be reached before fieldwork starts.

  3. 03

    Watch to the end, then answer

    Each viewer sees the whole ad before any questions, and the Next button only appears when it ends. Then they say which brand it was for, what it was telling them, how it made them feel and whether it makes them more likely to buy.

  4. 04

    Hear the reasons

    Open questions ask what people liked and what confused them. You can add a short video answer in which an AI interviewer asks up to three follow-up questions. Each clip is transcribed and checked for attention and lighting.

  5. 05

    Compare the cuts

    Each viewer sees one cut, so no version colors the view of the next. The report sets the cuts side by side with significance markers, and every result can be split by age, gender or any group you define.

  6. Then: decide

    After a reworked end card, before the media is booked.

    Plan your study

Which cut runs

Example: three cuts of a coffee ad

A US coffee chain has a 30-second film for its fall menu and two shorter edits of 15 and 6 seconds. Three matched groups of US adults aged 18-44 who buy coffee away from home each watch one version, then answer the same questions. The report shows which version people link to the chain most often, which message they play back, and where viewers lose the thread. The team picks a cut and reworks the end card before booking media.

  • 15-second cut

    Run it

    The most viewers name the chain and play back the message: 62% and 58%.

    Rework the end card before booking media.

  • 30-second cut

    Liked, less remembered

    The best liked, but viewers lose the thread in the middle.

    Keep it for places people choose to watch.

  • 6-second cut

    Brand gets lost

    Too quick: 12% name a rival brand.

    Show the brand in the first second.

Ask the studyAnswers from your own data

Did the 15-second cut do better with under-35s?

Yes. 66% of under-35s named the chain after the 15-second cut, against 51% after the 30-second cut. The gap is significant.

It checks its numbers against your data and draws the chart.

What you get

What you get

  • Output 01

    A scorecard per cut

    Brand recall, message, appeal and persuasion for every cut, with a marker where a gap between cuts is statistically significant.

  • Output 02

    Reasons in their words

    Likes, dislikes and the message people took away, grouped into themes with sentiment and example quotes.

  • Output 03

    Video reactions, if you add them

    Short video answers with transcripts and up to three AI follow-up questions each. Clips are for research use; showing one outside research needs the person's separate consent.

  • Output 04

    Results by audience group

    Compare up to five groups at once, such as under-35s and over-35s, or category buyers and non-buyers.

  • Output 05

    Answers to new questions

    Ask the AI analyst something like "Did the 15-second cut do better with under-35s?" and get an answer with a chart. It checks its numbers against your data.

  • Output 06

    Files to share

    The full report as a PowerPoint deck, an Excel workbook for your own tables, and raw CSV data.

FAQ

Questions, answered

Can I test a rough cut or an animatic?

Yes. Rough cuts and animatics are good enough to choose a route. Compare like with like, though: a finished film usually scores differently from a sketch, so do not set an animatic against a final edit.

Do you read people's faces to measure emotion?

No. People tell us how the ad made them feel by picking a face on a scale. We do not use their camera to judge their expressions.

Will you compare my ad with industry norms?

No. We do not have norm data yet, so we do not show norms. Your cuts are compared with each other, or with targets you set before fieldwork.

Does a must-watch view show whether people would notice the ad?

No. Because everyone watches to the end, the test tells you whether the ad is understood and linked to your brand, not whether it would stand out among other ads. For that, see social feed testing.

Can I run the test outside the US?

The video ad test runs with US audiences first. If you need another market, ask us and we will tell you what is possible.

Related

Goes well with

SolutionAudio ad testingHear how your audio ad landsSolutionStatic & poster testingSee what people notice first on your posterSolutionSocial feed testingWill your ad stop the scroll?OverviewAll solutionsEvery study we run, by the decision it supports.

Test your next ad before it airs

Tell us about the cuts, the audience and the decision you need to make. We will suggest the design and the viewers to recruit.

Talk to sales

Nearest office: Orlando, FL (corporate office)