Ad Concept Testing: Interview Guide and Questions
Use this guide to test 1 to 3 ad concepts, from storyboards to finished cuts, with your target audience. It checks recall, brand link and main message before it asks about feelings and fit.
- Length
- 20 min
- Questions
- 13
- People
- 12 to 18
- Sections
- 5
What this guide finds out
- What viewers remember from the ad without help, and in which order.
- If viewers link the ad to your brand or to a competitor.
- The main message that viewers take away, compared with the intended message.
- Which parts viewers find confusing, not believable or off-putting.
When to use it
- You have 2 or 3 creative routes and must choose one before production.
- You plan to spend a large budget on one ad and want to check the message first.
- An ad performs badly in market and you want to know what viewers took from it.
Who to talk to
People in the target audience of the campaign, with a quota for customers and for non-customers. Talk to 12 to 18 people. Each interview takes about 20 minutes.
The discussion guide
Each question has a probe note. The note tells the moderator, a person or an AI, what to do with the answer.
Warm-up
2 minLearn where the participant sees ads in this category.
Where do you usually see ads for [the category], for example on social media, TV or in the street?
ProbeRecord the main channel. Do not ask about any brand.
Exposure and recall
5 minCapture what viewers remember before any questions direct their attention.
I will now show you some content. Please watch it as you normally would.
ProbeShow the ad once, at normal speed. Place it between two neutral posts if the ad is for a feed. Do not say that it is an ad test.
What do you remember from what you just saw?
ProbeRecord items in the order they say them. Do not ask about scenes they do not mention.
Which company or brand was the ad for?
ProbeIf they name a competitor or do not know, record it as a brand link failure. Do not tell them the answer.
Message
6 minCompare the message viewers take away with the intended message.
In your own words, what was the ad trying to tell you?
ProbeWrite down their words exactly. If they describe the story only, ask what the story was meant to say.
Now watch it once more. Tell me about any part that is not clear.
ProbeShow the ad again. Ask what they thought was happening at that moment.
Is there anything in the ad that you do not believe?
ProbeAsk what would make it believable. Record the exact claim or scene.
Who do you think this ad is for?
ProbeAsk if they feel it speaks to them. If not, ask what would need to change.
Reaction
5 minLearn the feelings and actions the ad causes.
How did you feel while you watched it?
ProbeAsk at which moment the feeling changed, if it did. Accept "nothing" as a valid answer.
If you saw this ad in your feed tomorrow, what would you do?
ProbeRecord if they would skip, watch, search or click. Ask what made them choose that.
Is there anything in the ad that could annoy or offend someone you know?
ProbeAsk who and why. This framing lets people raise a concern without making it their own.
Comparison and close
2 minCompare concepts when the participant saw more than one.
Of the ads you saw today, which one would you remember next week, and why?
ProbeAsk this only if they saw more than one concept. Do not ask which one they like most.
Is there anything else about the ads that you want to tell me?
ProbeThank them and end the interview.
Example finding
Most viewers remembered the dog and the ending of the ad, but several linked it to a larger competitor.
“The one with the dog in the rain? That was the bank ad, the big one, I think. Or was it insurance?”
Mistakes to avoid
- Do not show the brand name or logo before the ad. You then cannot test if the ad builds a link to the brand.
- Do not show the ad many times before recall. One view is closer to what happens in a real feed.
- Do not ask the creative team's questions, such as "Is the music right?". Viewers react to the whole ad, not to parts.
- Do not show concepts in the same order to everyone. The first concept sets the standard for the rest.
Analyze the answers
Analysis templateThematic Analysis Codebook: Starter Codes for InterviewsUse this codebook as the first pass on interview transcripts about a task or a workflow. The 10 codes describe what people do and why, so themes come from behavior, not from opinions.Questions
Yes. Test a storyboard, an animatic or a rough cut. Tell participants that it is an early version so they do not judge the production quality.
Measure recall without help, the brand link, the main message, belief in the claims and the likely action. Liking alone does not predict how an ad works.
Yes. User Evaluation can show a video at a chosen question. It can also shuffle the order of what you show. The AI moderator only hears the participant, so ask them to describe what they saw.
Let the AI moderator ask these questions
The AI moderator interviews screened people by voice or text. Each claim in the study report links to its clip. The first 3 interviews are free.