Survey Design

Advertising and Brand Research Questions: Examples

Advertising and brand research questions with examples for ad pre-testing, brand awareness, consideration and brand health tracking, plus wording tips.

On this page
  1. What advertising research is
  2. What kinds of research questions there are
  3. Pre-testing an ad: example questions
  4. Brand health tracking: example questions
  5. Testing names, logos, and taglines
  6. Writing and running these tests
  7. Analyze and compare
  8. Common mistakes

Advertising research asks whether an ad is noticed, understood, liked, and believed, and whether it changes what people intend to do. Brand research asks what people think a brand stands for, how well they remember it, and how it compares with alternatives. The two use overlapping questions, but they serve different decisions. This guide gives example questions for each, explains how to word them, and covers how to compare results.

What advertising research is

Advertising research tests how well an advertisement works before or after it runs. Pre-testing shows people a draft ad or concept and asks what they take from it. Post-testing and tracking measure what people remember or think after the ad has run. Both help a team decide whether to change the creative, the message, or the media.

What kinds of research questions there are

Most marketing research questions fall into four groups, and each suggests a different kind of question:

  • Exploratory questions ask what is going on and what people think, such as “What comes to mind when you see this ad?” They usually call for open answers.
  • Descriptive questions measure how many people think or do something, such as “How many people recall seeing this ad?”
  • Comparative questions ask which option performs better, such as “Which of these two ads do you like more?”
  • Causal questions ask whether one thing changes another, such as whether seeing an ad changes intent. These need a causal research design with a comparison group, not just a single survey.

Pre-testing an ad: example questions

Show the ad as it will appear in the market, such as on a phone or on a screen, and keep the test to one ad or to a small set that you compare fairly.

Recall

Ask about recall before you show the brand, so that you do not prompt people.

  • Unaided: “Think about the last few ads you saw for [product category]. Which brands do you remember?” (open)
  • Aided: “Which of these brands do you remember seeing in an ad recently?” (list the brand and several others, select all that apply)

Unaided recall shows how easily people bring up the brand unprompted. Aided recall shows whether people recognize it when they see it. The two answer different questions, so report them separately.

Comprehension

  • “In your own words, what is this ad selling?” (open)
  • “Who do you think this product is for?” (open)

If many people give the wrong answer, the message is not clear, whatever the score on appeal.

Appeal and believability

  • “How much do you like this ad?” (Not at all to Very much)
  • “How much do you agree with this statement: this ad is believable.” (Strongly disagree to Strongly agree)
  • “What, if anything, would you change about this ad?” (open)

Avoid questions that assume a positive view, such as “How much did you enjoy this great ad?” The guide to leading and biased survey questions explains how to fix that kind of wording.

Intent (with caution)

  • “After seeing this ad, how likely are you to try [brand] in the next three months?” (Very unlikely to Very likely)

Stated intent tends to overstate what people do. Use it to compare ads with each other, not as a forecast of sales. The Likert scale guide explains how to set up the agreement and likelihood scales.

Brand health tracking: example questions

Tracking studies ask the same questions over time, so that changes can be measured. Keep the wording and the scales fixed between waves.

Awareness

  • “Which brands of [product category] can you name?” (unaided, open)
  • “Which of these brands have you heard of?” (aided, list of brands with several decoys)

Consideration and usage

  • “Which of these brands would you consider the next time you buy [product category]?” (select all that apply)
  • “Which of these brands have you used in the past 12 months?” (select all that apply)

Brand attributes

Brand attributes describe how people see the brand. Rate each one on the same scale.

  • “How much do you agree that [brand] is trustworthy?” (Strongly disagree to Strongly agree)
  • “How much do you agree that [brand] is different from other brands in this category?” (Strongly disagree to Strongly agree)
  • “What, if anything, makes [brand] different from other brands in this category?” (open)

Include a few negative statements, such as “[brand] seems outdated,” alongside the positive ones. Mixing them helps catch people who agree with everything.

Testing names, logos, and taglines

Before a launch, test the parts of the brand separately. Common questions include:

  • “How much do you like the name [name]?” (Not at all to Very much)
  • “How well does this tagline fit the product?” (Not at all to Very well)
  • “Please share what you like and dislike about this logo.” (open)

When you test several names, show them in a random order, so that the first name does not get extra attention. Ask about each name on the same scale. Be careful with names that sound too much like existing brands, because a survey can show liking without showing confusion. A short check of whether people can match the name to the product is often more useful.

Writing and running these tests

  • Use the same wording over time in tracking studies, so the changes you see are real.
  • Show the ad in its real format. An ad seen on a phone is not the same as one seen on a large screen.
  • Randomize the order of ads and brands when you compare them.
  • Keep the scales balanced, with equal positive and negative options.
  • Pilot the stimulus. Check that people can see the ad clearly and that the questions make sense. The pilot testing guide explains how.

Analyze and compare

Report the difference between the ad or brand and a benchmark, such as a competitor or your previous wave, not just the raw score. If you compare two groups, check that each has enough people to support the comparison. For small groups, a difference may be noise. Look at the spread of answers as well as the average, and report unaided and aided results separately.

Keep in mind what the survey can and cannot show. It can tell you how people react to an ad in a survey setting. It cannot show what they will do in the market once the ad runs against everything else competing for their attention.

Common mistakes

  • Prompting recall too early. Showing the brand before asking about unaided recall inflates the result.
  • Using one question for several things. “How much do you like the ad and believe it?” needs two questions.
  • Treating intent as a forecast. Use it to compare options only.
  • Changing wording between waves. Tracking results are only comparable if the questions stay the same.
  • Testing the audience you hope to reach, not the people who would see the ad. Match the sample to the real audience, as described in the segmentation guide.

For the wider design of a research project, including the technique and the sample, see the market research survey questions guide. For new products, the new product questionnaire guide covers concept testing in more depth.

The complete guide to survey design

Market Research Survey Questions, Instruments and Techniques
Read the guide

Keep exploring

Guides from other topics

All topics →