Creative Testing: How to Pre-Test Ads and Copy Before Launch

Creative testing is showing ads, scripts or copy to people in your target audience before launch and measuring four things: whether the work gets attention, whether people credit it to your brand, what message they take away, and whether it changes what they think or plan to do. A good test tells you which version to run and what to fix in it.

Branding, the second of those four, is often where creative falls down. In our Super Bowl ad testing study with 56 viewers, 59% said humor made an ad stick, but only 41% remembered an ad because the brand was part of the idea. Some viewers could retell a joke in detail and still not name who paid for it.

What creative testing should measure

A useful pre-test covers the full path from noticing an ad to acting on it.

Keep the order fixed, memory first and prompted questions last, so earlier questions don't feed answers to later ones. Our article on types of bias in research covers order effects and leading wording in more detail.

Branding and recall in ad testing

Enjoyment and brand credit came apart by 18 points in the Super Bowl study. When viewers did remember the brand, the ad usually made the product part of the joke, showed what the product does, or set the story somewhere that obviously belonged to that brand.

Celebrities showed the same split. 78% said the right celebrity makes an ad more memorable, yet 45% said the star had overshadowed the brand or its message. Ask viewers what the celebrity was selling. If they can't say, the creative is buying attention for the celebrity rather than for you.

Measuring message takeout and persuasion

Recall alone will not show you backlash in ad creative testing. In the Super Bowl study, 51% left the game with a better view of at least one brand and 32% left with a worse view of one, citing privacy worries, an off-putting tone or a celebrity who didn't fit.

Be realistic about behavior too. 67% of viewers took no action because of an ad, 23% searched online and only 6% bought something. A pre-test that asks only about purchase intent will overstate what the ad will do. Ask what people would search for next, so the landing page can answer it.

Qualitative vs quantitative creative testing

Scores pick the winner, and the reasons behind them tell you what to change.

Ask for both in the same session. A rating or forced choice comes first, then an open question about why, then a follow-up on whatever the person said. AI-moderated interviews make this practical at sample sizes that used to require a survey. On Listen, each participant rates the creative and then explains the rating out loud, and the answers are coded into themes you can count. For how to code those answers well, see our guide to qualitative data analysis, and for the wider trade-off, qualitative vs quantitative research.

How to set up an ad copy testing study

Our restaurant message testing study is a clean template for copy testing. We wrote 12 messages for a fast casual restaurant, three in each of four categories, and showed every message to 96 US diners one category at a time. For each message, diners rated its appeal, explained what worked or didn't in their own words, and said whether it made them more or less likely to visit.

Three design choices made the results useful:

• Several versions per idea. Two messages about saving money landed 23 points apart (82% vs. 59%). Testing one line per idea would have confused a weak line with a weak idea.

• A negative answer option. "Healthy Enough to Be Your Habit" made 14.6% of diners less likely to visit, the largest share in the study. A scale with no negative option would have hidden that.

• A why after every rating. Diners explained that the winner felt like clear value because it named a price and a rewards payoff, which tells the copywriter what to keep.

The same structure works for headlines, ad lines and scripts. Swap in your copy and keep the behavior question and the follow-up.

Testing AI-generated creative

AI-made ads need their own check, and in our data the quality of the finished ad mattered more to viewers than the fact that AI was used. In the Super Bowl study, after viewers were told one vodka ad was made entirely with AI, 57% called AI-made ads lazy or inauthentic, while 55% said knowing an ad used AI wouldn't change how they feel about it. Visuals that looked slightly off cost trust quickly. Viewers were more accepting when the work was polished and when synthetic people on screen were disclosed.

The stakes are higher in social commerce. In our TikTok Shop shopper insights study, 57% of shoppers rejected a product or seller when the content looked AI-generated or fake, and another 24% kept considering the product but checked it more carefully. Once something looked fake, shoppers started doubting the reviews and the seller as well. Put the finished cut in front of real people and ask whether anything felt wrong before you scale it.

Choosing a creative testing platform

Judge a creative testing platform on how well it answers the four core questions with your real audience. Look for recruiting that matches your buyers and screens out low-quality respondents, follow-up questions that probe each answer, open-ended answers coded into themes you can count, and every theme linked back to the participant's own words so the team can check it. Speed matters as well, because a test that returns after the media plan is locked can't change the ad.

Listen runs creative tests as AI-moderated video interviews. You add the ads or copy as stimulus, set the rating and follow-up questions, and get results in days with each number tied to the reasons behind it.

FAQ

What is creative testing?

Creative testing is the practice of showing ads, scripts or copy to your target audience before launch to measure attention, brand recall, message takeout and persuasion. The results show which version to run before the media money is spent.

What is the difference between ad testing and creative testing?

The terms are often used interchangeably. Ad testing usually refers to testing finished or near-finished ads, while creative testing also covers earlier work such as concepts, scripts, headlines and copy lines.

How do you test ad copy?

Write several versions of each idea, show each one to people in your target audience, and ask them to rate it, explain their rating in their own words and say whether it makes them more or less likely to act. Include a negative answer option so you can see which lines push people away.

What should an ad test measure besides recall?

Measure whether people credit the right brand, what message they take away, and whether the ad changes their opinion of the brand in either direction. In a Listen study of Super Bowl viewers, 32% came away with a worse view of at least one brand, which a recall-only test would miss.

How do you test AI-generated ads?

Show the finished cut to real people and ask whether anything looked off or felt wrong, then ask how it affected their trust in the brand. Viewers in Listen's Super Bowl study were more accepting of polished work and disclosed synthetic people than of visuals that looked slightly off.