Test video ad cuts with a consumer panel before client review
Two cuts, one client meeting, a split room. The five-step workflow that puts both cuts in front of a consumer panel and brings the report to the review link.
Two cuts, a client meeting next week, and a room that cannot agree. The editor likes the second opening, the producer likes the first, and the client has already asked which one you recommend. What wins in the edit bay and what lands with a stranger are not always the same thing, and a consumer panel can tell you which before anyone has to argue for it.
The situation
A recommendation with nothing behind it is one more opinion in the room. A written report from people who watched the cut cold, matched the buyer, and answered the one question the argument is really about is a different thing to bring to a meeting. The test does not replace the review. It gives the review a starting point that nobody in the room had a hand in.
Step one: finish both cuts to the same polish
Graded picture beats ungraded, a real mix beats temp music, and a final read beats scratch voice, for reasons that have nothing to do with the edit. If one cut is finished and the other is not, you already know the result. Bring both to the same level, match the length and the framing, and change one thing between them.
Step two: pick the one question
One primary question decides the verdict, and it is frozen when the test launches. That stops anyone, including you, from reading the results afterwards and promoting whichever number looks best. Choose the question you would defend in the meeting: does the opening hold, does the message land, would people pass it on. Supporting questions still come back as diagnostics, and the open answers come back as quotes.
Step three: choose the audience that matches the buyer
The general population is the base rate. If the argument is about how the piece reads to a particular viewer, shape the audience by age or gender so the people watching look like the people the work is for. A narrower audience costs more per person, and the wizard shows that as you choose.
Step four: set the size and read the quote
Participants are recruited from research panels and paid for their time, so the number of people is the biggest lever on the price. A short recruited test works out to about 4 credits per person, and one credit is one dollar: 24 people is around 100 credits, 48 around 200, 200 around 800. A paid test has a floor of 25 credits. A longer video or a narrower audience raises the per-person cost. The wizard shows the exact price before launch, and nothing is charged until you approve it.
Step five: launch and let it fill
Responses land while you get on with other work. The report is written when every version has reached the size committed at launch, and any seat that did not fill is refunded at settle.
What comes back
The report opens with The Bottom Line: the verdict in plain sentences and what it means. Then In Their Words, the participants' own comments, attributed and real. Then By the Numbers. It says "outperformed" only when both versions reached their committed size and drop-off was balanced; otherwise "directional advantage" or "too close to call", and a too-close result is information, not failure.
The report opens with the verdict in plain sentences and what it means for the cut you send.
Bringing it into the review
Send the client a review link for the cut the test favoured, with the report attached as the evidence. Keep the other cut as a version on the same asset so it is there if they ask. The client comments on the frame, approves the exact version, and never needs an account.
When not to bother
Skip the test when there is one cut and the only question is whether the client likes it. Skip it when the two cuts differ in five ways, because a win tells you the bundle won, not which change did the work. If you would rather ask your own list than a panel, send the test through a share link and it costs no credits.
Questions
Can I send both cuts to the client anyway?
Yes. Send the review link for the cut the test favoured, with the report attached, and keep the other cut as a version on the same asset so the client can open it for context. Both live in one place.
What if the report says no clear winner?
Then you have learned that the audience did not separate the two on the question you asked, which is real information for the room. The report says so plainly, and you can bring the quotes and the numbers to the meeting as the basis for choosing on other grounds.
Ready to try it?
Host video, collect frame-accurate review, and see who's watching. Free to start.
Start free