How apparel brands should route sizing questions: product page, fit tool, or AI shopping assistant?

Apparel brands should put recurring, answerable sizing facts on the product page; use a fit tool when a shopper’s measurements or preferences are needed to recommend a size; and use an AI shopping assistant for nuanced, product-specific questions spanning fit, comparison, and use case. The right choice comes from the question’s complexity, not from the novelty of the interface. Real shopper questions show which layer is missing.

Start with the question, not the channel

The best destination for a sizing question depends on what the shopper must know or do next. A question asking for a fact belongs in content; a question requiring personal inputs belongs in a fit tool; a question combining several products, preferences, or uncertainties is a strong assistant use case.

Collect questions from product-page search, customer support, reviews, returns and exchanges, chat transcripts, and sales conversations. Preserve the shopper’s wording. “Does this run small?” reveals a different problem from “What size should I get if I’m between sizes and want a relaxed fit?”

Classify each question on four dimensions:

Dimension	Question to ask	Likely destination
Repetition	Do many shoppers ask the same thing?	Product-page content
Personalization	Does the answer depend on body measurements, usual size, or preferred fit?	Fit tool
Complexity	Does the shopper need a recommendation across products or attributes?	AI shopping assistant
Confidence	Can the brand answer consistently from approved product data?	Any surface, with guardrails

A question can belong in more than one place. For example, a product page can state that a jacket has a relaxed cut, a fit tool can recommend a size based on measurements, and an assistant can explain how that jacket compares with a slimmer style.

Put stable sizing facts on the product page

Sizing and fit guidance belongs in product-page content when the answer is consistent for nearly every shopper and can be understood without a conversation. The product page should be the source of truth for fit language, garment measurements, size conversions, and the assumptions behind them.

Useful product-page content includes:

The garment’s stated fit, such as slim, regular, relaxed, oversized, or straight
Body measurements or “to fit” ranges for every available size
Product measurements where they help shoppers judge length, width, or proportion
Model height, worn size, and relevant fit context when the brand can maintain it accurately
Fabric stretch, structure, rise, inseam, sleeve length, and other attributes that change how the item wears
A clear explanation of how to measure and whether measurements describe the body or the garment
Local size equivalents for the markets the store serves

A static size chart cannot answer every fit question, but it prevents the fit tool or assistant from having to repeat basic facts. Nielsen Norman Group’s guidance on size guides and product measurements also distinguishes body dimensions from product measurements and recommends explaining how measurements were obtained. That distinction matters: a shopper cannot compare a body measurement directly with a garment measurement without context.

Move a question into the page when it is frequent, easy to answer, and relevant to most visitors. If shoppers repeatedly ask whether a style is cropped, whether the fabric stretches, or whether the same size differs from another cut, those answers should not be hidden behind a tool or assistant.

Use a fit tool for personal size recommendations

A fit tool belongs in the journey when the shopper needs a size recommendation based on personal information. It should ask only for inputs that materially affect the recommendation, such as body measurements, usual size in a comparable brand, height, or preference for a close or relaxed fit.

A fit tool is most useful for questions such as:

“Which size should I choose?”
“I’m between two sizes—what do you recommend?”
“I usually wear a medium in this type of top; what should I order here?”
“I want a fitted look. Should I size down?”

The tool needs reliable inputs from the catalog. At minimum, that means consistent size labels, measurement definitions, fit descriptors, and product-level differences. If one product’s “waist” means body-to-fit range and another’s means garment circumference, a polished interface will still produce confusing guidance.

Show the reasoning behind a recommendation. “Choose medium because your waist measurement falls within this range” is more useful than an unexplained size badge. Let shoppers edit their inputs, see the effect of a different fit preference, and return to the product page without losing their selection.

A fit tool is not a substitute for product information. It can recommend a size, but it cannot compensate for missing details about stretch, length, cut, or how measurements were taken. It also should not imply certainty where the brand’s data does not support it; a clear caveat and a route to human help are better than false precision.

Use an AI shopping assistant for nuanced questions

An AI shopping assistant belongs where shoppers need an explanation, comparison, recommendation, or follow-up rather than a single lookup. It is especially useful when a question combines fit with product choice: “I’m 5'4", usually a small, and want a loose layer for travel— which of these works best?”

Good assistant questions include:

“Does this run small compared with your other jackets?”
“Which jeans work for a curvier hip-to-waist ratio?”
“I like the fit of this shirt but need more room in the shoulders—what else should I consider?”
“What should I wear for a cold, wet commute if I dislike tight sleeves?”
“Can you compare the length and fit of these two styles?”

The assistant should answer from approved catalog data, fit guidance, reviews, policies, and product-page content—not from generic fashion assumptions. It should ask a focused follow-up only when the missing information changes the recommendation. If the shopper asks for a size but has not provided any useful context, the assistant can point to the fit tool or explain what information is needed.

Keep the assistant close to the product under consideration. Nielsen Norman Group’s ecommerce product-page guidance recommends anticipating product questions and making comparable information available across variants. Its chat guidance likewise supports making help available from product pages, where clarification is needed. The assistant should reduce navigation, not send a shopper away from the decision.

An assistant should also know when not to answer. Ambiguous product data, unusual sizing cases, medical or body-sensitive questions, and policy exceptions may require escalation. The brand should define what the assistant can claim, what it must qualify, and when it should direct the shopper to a person.

Turn shopper questions into a routing system

The practical answer is usually not product page versus fit tool versus assistant. Strong apparel journeys use all three, with each surface handling a different layer of uncertainty.

Use this routing rule:

Publish it when the question is frequent, stable, and broadly useful.
Calculate it when the answer depends on shopper-specific measurements or preferences.
Explain and compare it when the shopper is weighing products or expressing a nuanced need.
Escalate it when the approved data cannot support a confident answer.

Consider the question, “Should I size up?” The product page should first state the item’s fit and relevant measurements. A fit tool should use the shopper’s inputs to recommend a size. The assistant should handle the follow-up: why that recommendation differs from the shopper’s usual size, or which alternative has a roomier cut.

This layered approach also prevents a common failure: using a chat interface to conceal weak product content. If the same answer appears repeatedly in conversations, promote it to the product page and update the underlying fit data. If questions remain individualized after the page is improved, route them to the fit tool or assistant.

Measure whether the guidance resolves uncertainty

Measure the question’s outcome, not just the number of interactions. A useful review connects the original question to whether the shopper found an answer, used a recommendation, continued browsing, purchased, contacted support, or later requested an exchange or return.

Review questions by:

Product, category, and size range
New versus returning shoppers
Country or size-standard context
Fit concern, such as length, tightness, looseness, stretch, or proportions
Surface used: page content, fit tool, or assistant
Answer confidence and escalation rate

Look for different signals in each surface. Product-page improvements should make repeated questions less common. A fit tool should produce completed recommendations and reveal where shoppers abandon the input flow. An assistant should resolve multi-part questions, recommend relevant products, and expose unanswered patterns.

Anagram’s homepage describes this as a loop: its Site Agent helps shoppers with conversational support while they compare and decide, while the company’s insights show what shoppers care about and where conversion friction appears. That makes it a relevant layer for capturing the real wording behind apparel fit questions and deciding which recurring answers should become page content, fit-tool logic, or assistant guidance.

Review the question set regularly. The goal is not to send every question to the most sophisticated tool. The goal is to make the simplest reliable answer available at the moment a shopper needs it, then use the remaining questions to improve the next version of the experience.