A quiz is a salesperson that scales
In a store, an associate asks two or three questions and hands you the right product. Online that conversation does not happen, so the undecided customer does the worst thing a customer can do. They open six tabs, compare, get tired, and leave. A product quiz is that associate, running on every session at once.
The quiz does two jobs at the same time. It moves an undecided shopper to a confident purchase, and it captures an email tied to everything the customer just told you about themselves. One surface, two of the highest-value outcomes on your store.
When a quiz earns its place
Quizzes pay off when the customer cannot easily self-select. Regimen-based catalogs like skincare and supplements, variant-heavy lines, anything sold by skin type, goal, or use case. If a new visitor has to read three product pages to figure out which one is theirs, a quiz will out-convert the catalog.
Skip the quiz when you sell one hero product, or when the choice is obvious at a glance. A quiz in front of a single-SKU store is a speed bump. The test is simple: if your support team answers the same which-one-is-right-for-me question all day, that question is your quiz.
Question design: fewer, sharper
Four to six questions, never more. Every question you add costs you completions, so each one has to earn its place by changing the recommendation. If two answers lead to the same product, the question was decoration. Cut it.
Map each question to a real axis of the recommendation. Skin concern, not favorite color. Show a progress indicator so the customer knows the end is near, write the options in their language, and let them tap an image rather than read a paragraph. The quiz should feel like being understood, not interrogated.
The recommendation: one confident answer
End on a single recommended product, or a tight set of two or three, never a wall. The customer took the quiz to be told what to get. A result page that returns ten options has handed the decision back and wasted the trust the quiz just built.
Show your work. A one-line reason for each pick, drawn from the answers they gave, turns a recommendation into advice. Because you said your skin is dry and reactive, here is the fragrance-free option. That sentence is the difference between a result that feels personal and one that feels like a coupon.
Capture the email without killing completion
Ask for the email after the customer has felt the value, never before. Gate the result behind an email field and a chunk of people abandon at the last step, taking their answers with them. Show the recommendation first, then offer to send it along with a reason to want it, a starter discount or the routine in writing.
Treat the email as optional and you will capture more usable addresses than a hard gate collects, because the ones you get actually want to hear from you. Completion rate and capture rate pull against each other. Tune the trade with the gate placement, and watch downstream conversion, not just the raw opt-in count.
Close the loop with retention
A quiz that only recommends is half a tool. The other half is the data. Every answer is a segmentation signal, so pipe the responses and the email straight into Klaviyo and tag the profile with the result. Now your flows know skin type, goal, and the exact product you recommended.
That turns a one-time quiz into the top of a retention engine. The welcome flow references their result. The replenishment reminder fires on their product's reorder cycle. The customer who spent thirty seconds answering six questions becomes the easiest customer you have to bring back, because you already know what they want.