How should we collect zero-party data such as quiz answers and preferences?
Our answer
Ask for preferences only where the answer changes what the shopper gets next. Start a quiz at 3 to 5 questions, ask one or two at each other touchpoint, and store every answer on the shared customer profile.
- Confidence
- Medium
- Strongest proof
- 1 controlled test
- 3 official rules, 4 practitioners, 4 opinions or claims
- Sources
- 11 sources
- 12 statements
- K
CLASO
New here? Take the 1-minute tour
Why this is the answer
3 reasons and our judgmentEvery question needs a use before it goes live. Darkroom (paid media agency) ships no question without an owner and a named segment or flow. Its quizzes run 3 to 5 questions, and the answers change the product page and welcome path.
Start small and ask a little at a time. Charle (Shopify agency) starts with a quiz and a sign-up form that adds one or two preference questions. It routes every answer into Shopify metafields and the email platform.
Answers only stick when the shopper is known. Klaviyo's documentation says form answers become profile properties only when the visitor gives an email or phone number or is recognized. Custom properties can then sync to Shopify customer metafields.
Our judgment
- Start a quiz at 3 to 5 questions and keep every other ask to one or two. That follows Darkroom's and Sparq's quizzes of three to five questions, LTV.ai's limit of 2 to 3 questions per collection moment and Charle's one or two preference questions on the sign-up form.
- Keep a longer quiz only if it wins on revenue per visitor in a test. Evolve Media runs 4 to 7 questions, and six more questions won in one brand's A/B test reported on Scalability School.
- Offer the preference center beside the unsubscribe link rather than in its place. Preferences should never stand between a reader and leaving, as Unsubscribes and spam complaints sets out.
- Keep each answer as a profile property in the email platform synced to the Shopify customer record, not only inside the quiz app. That follows Charle's advice to route every answer into Shopify metafields and Klaviyo's documented sync of custom properties to Shopify metafields.
- Judge each question by revenue per recipient against a control, as Darkroom does. Read Charle's comparison of customers who shared preferences with those who did not as a lean only.
Our judgment, built from Darkroom, Sparq, LTV.ai, Charle, Evolve Media, Scalability School, Litmus and Klaviyo: our reading of these sources, not a finding from any one of them.
What the sources say
12 statements from 11 sourcesKlaviyo says it can sync custom profile properties to Shopify as customer metafields, up to Shopify's 250 definitions per object, with matching data types. Deleting a profile in one doesn't delete it in the other.
“The number of metafield definitions per object is limited to 250 in Shopify, so you may only sync custom properties up to that limit.”
Klaviyo says sign-up form answers save to a profile only if the visitor gives an email or phone or is already known, a new answer replaces the old one, and separate forms need unique property names.
“If someone answers the same question again, the new value replaces the previous value for that property.”
Brad Ploch says a quiz page for a single-product brand took Meta spend like other landing pages, and split tests improved it: six problem-focused questions doubled revenue per visitor. Sample sizes are not given.
“sit in the problem longer”
Darkroom, a retention agency, says ask for zero-party data only where you will act on it, one question per interaction, with quizzes of three to five questions each mapped to a segment.
“Progressive profiling means asking one question per interaction across the customer lifecycle instead of fourteen fields at signup.”
Litmus relays MailerLite data that unsubscribe rates more than doubled from 2024 to 2025, largely from Gmail changes, and an Oracle finding that a snooze option cut unsubscribes 82%, method not given.
“senders who added a snooze option to their email preference centers saw unsubscribes drop by an average of 82%”
Charle, a Shopify agency, advises collecting customer preferences through a quiz and a sign-up form with one or two questions, routed into Shopify metafields and your CRM, and dropping unused questions.
“Asking a customer their preference and then ignoring it is worse than never asking.”
Evolve Media, which builds quiz funnels, says from 40+ client builds that a 4 to 7 question quiz with email capture before the result captures 3 to 5x more emails than a static form. No method shown.
“4-7 questions. 30-50% completion rate. Email capture immediately before the result reveal.”
Okendo, which sells quiz software, says a reward can lift email capture in quizzes, but requiring an email likely raises drop-off and invites fake addresses.
“requiring email is likely to increase drop-off rate and reduce overall completions and conversions.”
Omnisend says ask for the email after a quiz to unlock results, not before, and give away prizes tied to your products, since generic prizes draw freebie hunters who unsubscribe.
“Visitors will not start a quiz they know is just a lead capture.”
Sparq, which sells Shopify search and filters, suggests a product quiz of three to five questions mapped to filter attributes and a two-question post-purchase survey. Its 20 to 30% conversion claim is uncited.
“A two-question post-purchase survey on your order confirmation page is the easiest zero-party data collection you can implement.”
LTV.ai, which sells email personalization, advises asking no more than 2 to 3 preference questions at a time and starting small brands on the one attribute that matters most. Its unsubscribe claim cites no source.
“Keep surveys to 3 questions maximum.”
Shopify Analytics can now group and filter by metafields on products, variants, customers and orders, such as material or loyalty tier, once each definition is switched on.
“You can now group and filter by metafields on products, variants, customers, and orders, for example by material, loyalty tier, or ingredients”
In the full answer
For readers who sign in- What to do this week3 steps
- The numbers behind it2 figures from cited sources
- Where each source standsA chart placing each source on the question
- The strongest disagreementThe best case against the answer, from a cited source
- How sure we areWhat its confidence rests on, and what would change the answer