How to write Product Advisor questions that convert

The first thing most teams do when a Product Advisor under-performs is open the scoring sheet. They re-weight an attribute. They add a tie-breaker. They argue about whether “sensitive skin” should count for two points or three.

Almost none of that moves the number.

The thing that moves the number is the question text. Most Product Advisor drop-off is a phrasing problem. A shopper who hesitates on question two is not confused by your weights — they are confused by what you are asking and why.

Write the question a salesperson would ask

Stand on the shop floor for an afternoon and listen to how a good associate opens a conversation. They never say “Please select your skin concern from the following list.” They say “What's been bothering you lately?”

The first version is a form field. The second is a conversation. The shopper answers the second one in half a second because they already know the answer — they were thinking about it on the way in.

The test for every question in your Product Advisor: would a human ever ask this out loud, in those words, to a stranger? If not, rewrite it until they would.

Ask one thing at a time

The most common drop-off culprit is a question that is secretly two questions. “What's your skin type and main concern?” is two questions. “Who is this for and what's the occasion?” is two questions.

Shoppers stall on these because answering means choosing which half of the question to prioritise. Split them. Two short questions almost always out-convert one compound question, even though it looks like more work on paper.

Order matters more than length

A four-question Product Advisor in the wrong order will drop more shoppers than a six-question one in the right order. The rule of thumb:

1. Easiest answer first. The opening question should be one a shopper can answer without thinking. “Who is this for?” with three big tappable options beats “What's your budget?” every time. The first answer is a commitment device — once a shopper has tapped one button, the next four feel free.
2. Biggest disqualifier next. Now spend the engagement you just earned. Budget, allergen, size, climate — whatever cuts the catalogue in half in one tap. This is the question that makes the recommendation feel personal later.
3. Taste or preference last. Colour, flavour, style — the question that picks the winner from the survivors. Put this near the end where the shopper is already invested.

If a question is hard and not a strong disqualifier, it probably doesn't belong in the Product Advisor at all. Move it to the product page.

Show progress, always

A Product Advisor without a progress indicator feels infinite. Even three questions feels infinite if a shopper can't see the end. A simple “Question 2 of 4” or a thin progress bar is worth more than any copy change you will make this quarter.

The reason is simple: people will finish things they can see the end of. The same shopper who would abandon an open-ended experience will push through a four-of-four because the finish line is visible.

Make options mutually exclusive — or explicitly not

If two of your answer options could plausibly be true for the same shopper, you have a phrasing bug. Either rewrite so only one applies, or switch to multi-select and say so on the screen.

A shopper staring at “Dry skin” and “Sensitive skin” when their skin is both will either pick one at random or close the tab. Neither is the answer you want. “Pick all that apply” in small text under the question is one of the highest-leverage twelve characters you will ever ship.

Keep answers to four, maybe five

More than five options on a single screen and you have rebuilt the catalogue inside the Product Advisor. The whole point was to narrow choice; a question with nine answers does the opposite. If you genuinely need nine, that is a sign the question should be two questions — a broad one followed by a narrower one only the matching shoppers see.

Mobile is the forcing function here. Anything that requires a scroll to see all the options will lose roughly the share of shoppers whose answer was below the fold.

Watch the data, not the opinions

Once the Product Advisor is live, the drop-off chart will tell you exactly which question is wrong. The single steepest drop is almost always the one to rewrite first — not the one your team argues about most in the Slack thread.

When you do rewrite, change one thing at a time. Tweak the question wording or the option labels or the order — never all three in the same week. Otherwise you learn nothing from the next week's numbers.

And give each change at least a few hundred completions before you judge it. Small samples lie loudly.

The shortest checklist that works

Before you ship a Product Advisor, read each question out loud and ask:

• Would a human actually say this?
• Is it asking exactly one thing?
• Can the shopper answer in under five seconds?
• Are the options mutually exclusive — or clearly multi-select?
• Does the order go easy → disqualifying → taste?
• Can the shopper see how many questions are left?

Six checks. Most under-performing Product Advisors fail at least three of them. Fix the phrasing and the scoring will look a lot smarter than it actually is.