Purchase intent (PI) questions are blunt metrics that are known to have a low correlation to actual purchase behavior. However, researchers generally don’t use them solely as indicators of determining the extent to which people are likely to buy something. Instead, these questions help researchers compare brand and product preferences, help product developers understand customer needs, help marketers understand demand, and help sales professionals know where to focus their efforts. They serve a far greater purpose than just estimating purchase intent.

The precise phrasing of PI questions depends on the product and how often or when the product is used or purchased but there are many styles and formats to choose from, especially for researchers working with Fast-Moving-Consumer-Goods (FMCG) like personal care, household, OTC products, pet care, or food and beverage items. Let’s review some of the good, bad, and great ways we can write, design, and program purchase intent questions in marketing research questionnaires.

Boring, Biased Binary Radio Buttons

The simplest way to ask about purchase intent is unfortunately the most biased and least recommended format. By formatting PI as a binary, yes/no question, it becomes a leading question that entices people to “help” the researcher by saying “yes.” As much as possible, avoid binary options for PI questions.

1. The next time you shop for detergent, will you buy this brand?

• Yes

• No

‍

Forced Choice Single-Answer

Deservedly so, participants generally don’t like forced choice questions as they don’t allow people to speak their truth. Perhaps they would prefer to go without rather than buy any of them. Or, maybe they would make their own product or borrow some from a friend. However, this question could be appropriate in rare scenarios where there are truly no alternatives apart from the mentioned brands.

2. The next time you shop for detergent, which ONE of these brands will you buy?

• Brand A

• Brand B

• Brand C

• Brand D

‍

Multiple-Brand Multiple Choice

Alternatively, consider a multi-brand multiple choice with an opt-out answer option (e.g., Don’t know, Other, None of the above, Won’t buy any). This will appease people who wouldn’t buy any of the brands listed perhaps because they use a cleaning service, someone else does the buying, or they use homemade alternatives. And, it disguises the “correct” answer so people don’t feel like they should “help” the researcher.

3. The next time you shop for detergent, which of these brands will you buy?

• Brand A

• Brand B

• Brand C

• Brand D

• Opt-out

‍

Single-Purchase Package Counts

To switch things up, ask about the volume of purchase rather than simply which brands will be purchased. This works particularly well when the product category tends to be purchased in volume or as multiples. A purchase of one could be considered brand trial whereas a purchase of 2 or more could be considered as brand trust. This question could be formatted as one brand in a single answer question or several brands in a grid.

4. The next time you shop for detergent, how many of each of these brands will you probably buy?

5. The next time you shop for detergent, how many of this brand will you probably buy?

• 0

• 1

• 2

• 3

• 4

• 5 or more

‍

Multiple-Purchase Package Counts

If a product gets purchased frequently over a reasonably short time frame, it might be appropriate to ask not only about the next purchase, but also about the next five or ten purchases. As before, a single package could indicate trial whereas multiple packages could indicate trust. This question could work as a single answer or a grid question.

6. Thinking about the next ten packages of detergent you will buy, how many will be each of these brands?

7. Thinking about the next five packages of detergent you will buy, how many will be this brand?

• 0

• 1

• 2 or 3

• 4 or 5

• 6 or 7

• 8 or 9

• 10  

‍

Single-Brand Balanced Ratings

Rating questions are quite common but also very boring. They are best suited for when you need to compare results to previous questionnaires that used the same format. Typically, these questions are balanced meaning they include a neutral option in the center, flanked by an equal number of negative and positive choices. To prevent bias and to create meaningful comparisons, rating scale questions can also be designed as grid questions that include several brands.

8. If these detergents were on the shelf today, how likely are you to buy each of them?

9. If this detergent was on the shelf today, how likely are you to buy it?

• Very likely

• Somewhat likely

• Neutral

• Somewhat unlikely

• Very unlikely

‍

Single-Brand Unbalanced Ratings

You might instead consider an unbalanced scale when it is extremely unlikely that people will choose one end of the scale, e.g., a category is loved by 90% of the market. In such cases, the answer options won’t include a neutral middle point. Instead, they will include a range of options from high to low. This can help to spread out the answers.

10. On a scale from Extremely interested to Not interested at all, how interested are you in buying each of these brands?

11. On a scale from Extremely interested to Not interested at all, how interested are you in buying this brand?

• Extremely interested

• Very interested

• Somewhat interested  

• Slightly interested

• Not interested at all

‍

NPS Style

In general, scales that use more than 5 or 7 points are less desirable as they create false precision. But, if they’re asked alongside a couple other NPS style questions, using matched question formats simplifies the participant experience.

12. On a scale from 0 to 10, how likely are you to buy this brand the next time you need to buy detergent?

‍

Multiple-Brand Rankings

With ranking questions, we start getting a little more creative. These question formats can be set up in a variety of ways such as having participants type numbers beside brands, click arrows to move brands up and down, or drag and drop brand logos into images of shopping carts. Look through your survey tool to see what creative options are available to you.

13. The next time you shop for detergent, which ONE of these brands would you choose first? If that brand wasn’t available, which brand would you choose second and then third?

‍

Timer Questions

Fans of Daniel Kahneman’s System 1 and 2 Theory will appreciate this technique. We know purchase behaviors aren’t completely logical nor critical and question types that rely on speed may trigger more intuitive, emotional thinking. The theory is that asking people to work as quickly as possible will reveal their true opinions. If you’re interested in leveraging speed, make sure to turn on time recordings in your survey tool as those data may not automatically be added to your dataset.

14. When you’re ready, I will show you five brands. As fast as you can, click on the ONE brand you are most likely to buy. If you’re ready to begin, click on GO!

15. When you’re ready, I will show you five brands. As fast as you can, sort the brands from most likely to buy to least likely to buy. If you’re ready to begin, click on GO!

‍

Let People Speak Their Truth with Single or Multiple Text Boxes and Open Ends

Historically, open-ended questions were time-consuming and expensive to code. Today, however, the availability of AI and automated coding makes them a great option for encouraging people to share their personal truths. Open-ended questions allow people to share brands and details that could never be part of traditional questions such as locally-crafted brands, international brands, or homemade products.  

16. What brand of detergent will you buy next time? _________________________

‍

Get Creative

So far, we’ve considered different formats or styles of questions but we can do one better by also using engaging and interesting wording. By injecting some creativity, we can encourage participants to slow down and think more carefully. If you’re launching a questionnaire that will only be used once or twice, these kinds of questions are great options.

17. If you had to wait in line to buy this brand of detergent, how many minutes would you wait?

18. If you had three seconds to run down the store aisle and choose a detergent, which ONE of these brands would you buy?

19. If you could only have one brand of detergent for the rest of your life, which brand would you buy?

20. If this brand of detergent were on sale for 24 hours, would you click on 'buy' now or wait for another brand?

21. If a stranger secretly put this brand of detergent in your shopping cart, would you keep it or remove it?

22. If your socks could talk, which brand of detergent would they beg you to buy?

23. You’re trapped in a room with your gym clothes after a week of workouts. Which brand of detergent would you buy to wash them?

24. If your worst enemy was excited about this brand of detergent, would you buy it?

25. If aliens landed on earth and demanded the best detergent, which brand would you buy for them?

‍

Summary

It’s tempting to grab old questionnaires and copy every relevant question word for word. Not only does it speed up the process, it creates consistency among historical and fresh data. But consistency isn’t always necessary and speed doesn’t always generate the level of quality that is essential for understanding consumers’ current needs and desires.

When the opportunity arises, make the effort to get creative. Try a different question format. Try more interesting and engaging wording. Your research participants will thank you with more thoughtful and personally relevant answers.

If you’d like to test out different types of engaging question formats, please get in touch with one of our survey experts. We’d love to show you the full range of options you can choose from!

‍