Marketing Automation

Product Quizzes: Collecting Zero-Party Data That Improves Recommendations

How to build a product quiz that collects useful zero-party data, stores answers so flows can use them, and proves its value beyond the quiz completion rate.

TL;DR
  • A product quiz is worth building when shoppers struggle to choose, and when every answer changes either the recommendation or the follow-up messaging. If an answer changes nothing, cut the question.
  • Store answers as structured profile properties with fixed values, so your email and SMS flows, segments and on-site recommendations can read them. Free-text answers and inconsistent labels make the data useless.
  • Measure the quiz on revenue and margin from quiz takers, compared against a fair baseline, not on completion rate. Quiz takers are already more engaged, so a holdout gives the honest answer.

What is a product quiz and what zero-party data should it collect?

A product quiz asks shoppers a few questions and recommends products based on their answers. It works best for catalogs where the right choice depends on the person: skin type, hair type, fit, taste, pet age, coffee brewing method or experience level.

The answers are zero-party data: information a customer intentionally shares with you. That is different from first-party data, which you observe from behavior such as page views and orders. Behavior tells you what someone did. Quiz answers tell you what they want and why.

That combination is what makes the quiz valuable beyond the first sale. A store that knows a customer’s concern, routine and budget can send messages that fit, instead of guessing from the last product viewed.

Collect three kinds of answers:

  • Fit answers that determine the product: type, size, intensity, format.
  • Context answers that shape messaging: goal, experience level, who it is for.
  • Timing answers that shape when to follow up: how often they use the product, when they expect to run out, an event date.

Which questions should a product quiz ask?

Every question should earn its place by changing the recommendation, the segment or the timing. Each extra question also gives shoppers one more chance to quit, so keep the quiz as short as the recommendation logic allows, and put the easiest questions first.

Question type Example Changes the recommendation Used in messaging Store as
Primary need “What is your main skin concern?” Yes Yes, content by concern quiz_concern
Type or fit “How would you describe your skin?” Yes Sometimes quiz_skin_type
Experience level “How long have you used a serum?” Sometimes Yes, beginner or expert content quiz_experience
Usage pattern “How often do you use this product?” No Yes, reorder timing quiz_usage_frequency
Budget or format “Do you prefer a full routine or one product?” Yes Yes, bundle vs single offers quiz_routine_size
Who it is for “Is this for you or a gift?” Sometimes Yes, gifting flows quiz_recipient

Use multiple choice, not free text. Fixed answer options give you clean values to segment on. Use a free-text field only where you will actually read it, such as an optional “anything else we should know?” at the end.

Avoid questions about sensitive categories, such as health conditions, unless you have a clear reason, a clear privacy notice and advice from counsel on how that data can be stored and used.

Where should the email capture go in the quiz?

You have two main options.

  • Gate the results. Ask for an email before showing recommendations. You capture more addresses, but some shoppers quit at the gate, and some enter throwaway emails.
  • Offer, do not require. Show the results, then offer to email them along with a routine guide or a first-order offer. Fewer captures, but they are more willing.

There is no single right answer. Test both on your traffic and compare revenue from quiz takers over 30 to 60 days, not the capture rate alone.

Whichever you choose, keep consent clean:

  • Email consent and SMS consent should be separate choices. SMS marketing in the U.S. requires express written consent under the TCPA, with clear disclosure. Our guide to SMS compliance covers the basics; confirm your setup with counsel.
  • Tell shoppers how you will use their answers, in plain language near the capture field.
  • If deliverability matters to your list, consider double opt-in for quiz signups, especially when the quiz is promoted through paid traffic.

How do you store quiz answers so automation can use them?

This is where most quizzes fail. The quiz tool collects answers, but they never reach the email service provider in a usable form, so flows ignore them.

Set up four things:

  1. Profile properties with fixed names and values. Use a prefix (quiz_), lowercase names and a fixed list of values. “oily” and “Oily skin” are two different segments to a machine.
  2. A completion event. Send an event such as “Quiz Completed” with the answers and recommended products as properties. Events can trigger flows; properties alone usually cannot.
  3. A quiz version and date. Store quiz_version and quiz_completed_at. When you change the questions, you can still tell which answers came from which version, and you know how fresh the data is.
  4. Analytics events. Track quiz start, each step and completion with event tracking in GA4, so you can see where shoppers quit.

Check your quiz app’s integration before you build. Confirm which fields it sends to your ESP, whether it sends events or only properties, and whether answers update or overwrite when a customer retakes the quiz.

If you use our Value × Intent (RFMB) model, quiz answers sit next to it as a third layer. RFMB tells you how valuable a customer is and where they are in the buying stage. Quiz answers tell you what they need. Together they decide both whether to send and what to say.

How do you turn quiz answers into flows?

Quiz data pays off in four places.

The quiz results flow

Trigger a flow from the completion event. The first email repeats the recommendations, explains why each product fits the answers given, and links straight to the products. Follow-ups branch by the primary need: a shopper concerned with dryness gets content and reviews about dryness, not a generic bestseller email. Exit the flow when the shopper buys.

Segmented campaigns

Use quiz properties in audience segmentation for campaigns. A launch email for a new oil-control product goes first to the “oily” segment. Our guide to email segmentation covers how to combine segments like this with engagement and purchase history.

Timing and replenishment

A usage frequency answer gives you a starting estimate for when a customer will run out. Use it for the first reorder reminder, then switch to the customer’s actual reorder timing once you have it.

Post-purchase and on-site

After the first order, use the quiz answers in onboarding emails (“how to use your serum for dry skin”) and in on-site recommendations when the customer returns. This is personalization based on what the customer told you, which is easier to explain and less likely to feel intrusive than personalization based on tracking alone.

Worked example: from answers to a recommendation map

All numbers below are illustrative.

A skincare store’s quiz asks three fit questions: concern (dryness, breakouts, dullness), skin type (dry, oily, combination) and routine size (one product, full routine). That gives 18 answer combinations. Instead of writing 18 recommendations, the team builds a map:

Concern Routine size Recommendation Follow-up content
Dryness One product Hydrating serum Dryness tips, serum reviews
Dryness Full routine Hydration bundle Routine order, bundle reviews
Breakouts One product Clarifying serum Breakout tips, ingredient guide
Breakouts Full routine Clarifying bundle Routine order, before and after guidance
Dullness One product Vitamin C serum Brightening tips
Dullness Full routine Brightening bundle Routine order

Skin type then adjusts only the moisturizer choice inside each bundle. Six core outcomes are easy to maintain, merchandise and measure.

The funnel is measured monthly: 10,000 quiz starts, 6,000 completions, 3,000 email captures and 450 orders from quiz takers within 30 days. The team tracks each step separately. If completions drop when a question is added, the question costs more than it returns.

How do you measure whether the quiz pays off?

Quiz takers convert better than average visitors almost by definition, because people who spend two minutes answering questions are already interested. Comparing them with all visitors overstates the quiz’s value.

Use better comparisons:

  • Holdout. Show the quiz entry point to most visitors and hide it from a random holdout group. Compare revenue per visitor between the groups, not between takers and non-takers.
  • Flow performance. Compare the revenue per recipient of the quiz results flow with your other signup flows, such as the pop-up welcome series.
  • Fit quality. Compare return rate and repeat purchase rate for orders of quiz-recommended products against other first orders. A quiz that recommends better should reduce “not right for me” returns.
  • Step drop-off. Use GA4 events to find the question where most people quit, then shorten or reword it.

Checklist: product quiz setup

  • [ ] Every question changes the recommendation, a segment or the follow-up timing.
  • [ ] Answers are multiple choice with fixed values.
  • [ ] Profile properties use a consistent prefix and naming.
  • [ ] A completion event with answers and recommended products reaches the ESP.
  • [ ] Quiz version and completion date are stored.
  • [ ] Email and SMS consent are separate, with clear disclosure.
  • [ ] A results flow triggers on completion and branches by primary need.
  • [ ] Quiz properties are available as segments for campaigns.
  • [ ] GA4 tracks quiz start, each step and completion.
  • [ ] Results are measured with a holdout or a fair comparison, not against all visitors.

FAQ

How many questions should a product quiz have?

As few as your recommendation logic needs. Map the questions that change the outcome first, add one or two that improve messaging, and watch step drop-off. If a question causes many shoppers to quit and changes nothing important, remove it.

Should I offer a discount at the end of the quiz?

Only if you would offer it to a new subscriber anyway. Many quizzes convert without one because the recommendation itself is the value. Test a version without the offer before you make a discount the default.

Can I use quiz answers in paid ads?

You can build audiences from customers who consented to marketing, within each platform’s rules. Avoid using sensitive answers for targeting, keep your privacy notice accurate, and confirm with counsel where the data includes anything personal or health-related.

Turn this into a plan for your store

The Free Profit Audit scans your store, sizes your biggest profit leaks and builds a 90-day plan.

Free Profit Audit