blog / Ecommerce

How to help customers choose the right size on WhatsApp without guessing

heysteff is the AI platform for customer service and sales over WhatsApp, Instagram, Messenger, Gmail and Shopify; Steff is the AI agent that runs it. For apparel and footwear stores, helping customers choose the right size on WhatsApp is one of the most common conversations and one of the most decisive: answer late or vaguely and the shopper leaves, answer badly and they buy, then ask for an exchange. This guide covers what to ask, which information to have ready and when a person should take over.

Why the size question on WhatsApp decides the sale

The size question decides the sale because it is the last doubt before checkout: someone asking which size fits has already picked the product and only needs confidence. By size question we mean any message where a shopper asks for help choosing between sizes, whether they share measurements, the size they wear in another brand, or a description of their build ("I have broad shoulders", "my feet are wide").

The underlying problem is that sizing information on most stores falls short. Baymard's usability research found that more than 80% of apparel sites don't provide sufficient sizing information, and one of its recommendations is to put a customer service link right inside the size guide. For a growing number of stores, that link is a chat window, and on WhatsApp-first markets it is simply the store's number.

That cuts two ways. The bad news: every size question that waits hours for a reply is a purchase cooling off. The good news: a well-handled conversation closes the sale and prevents the exchange that would follow, which in apparel costs shipping, staff time and margin.

What to ask before recommending a size

Before recommending a size, ask a few specific questions: who it's for, what size they wear today in a similar item, and how they like it to fit. Those three answers solve most cases without asking anyone to find a tape measure.

Who it's for. A gift changes everything: the buyer can't try anything on and often doesn't know the exact size. Recommend the safest option and mention up front how exchanges work.

What they wear in a similar item. "I wear a medium in cotton tees" or "I'm a 7.5 in running shoes" is far more useful than a generic size, because it compares like with like. If they name a brand and your team knows how it runs against yours, that's gold.

How they like the fit. Snug, regular or relaxed. The same person may need two different sizes depending on use: a training tee isn't chosen the same way as an oversized hoodie.

Only if that isn't enough does it make sense to ask for measurements: chest, waist and hips for clothing, or foot length for shoes. Asking for measurements first scares off someone typing on their phone during a commute; asking at the end, when there's a real doubt between two sizes, feels like service.

The fit data your team or AI needs on hand

Answering size questions well takes product-level data, not just a generic chart: actual garment measurements, the cut, and how the fabric behaves. A garment measurement chart shows the dimensions of the finished item (width, length, sleeve), while a body measurement chart shows which body fits each size. Ideally you have both.

Beyond the charts, write down three things that usually live only in your most experienced rep's head: whether the style runs large, small or true to size; whether the fabric stretches or shrinks after washing; and what size the model in the photos is wearing and how tall they are. That last one is the most asked and the least published.

If you use an AI agent, this information is what separates a useful answer from a made-up one. An agent that only sees the product name and its list of variants can't know that one pair of jeans runs small. That's why the AI shopping assistant needs fit notes per style or product line, written in the words your customers actually use.

A quick shortcut: read your last fifty size questions. The same five or six doubts repeat, and each one can become a fit note or one more line on the product page.

How to answer: recommend a size, don't paste the chart

The right answer to a size question is a specific recommendation with a reason, not a link to the chart. If the shopper chose to message you, the chart already failed them; sending it again hands the work back.

Here's an example. A sportswear store gets: "Hi, I usually wear a medium but I've got a broad back, what size should I get in the compression shirt?" A weak reply is "Here's our size guide." A useful reply is: "I'd go with the large. It's a compression fit and runs snug, so with a broad back the medium will feel tight across the shoulders. If you like it really fitted the medium works too, but the large is the safer pick. Black or grey?"

That reply does four things: it recommends a size, explains why using a fact about the garment, acknowledges the alternative and returns to the sale with a closing question. It's the same logic behind recommending products to undecided shoppers: fewer options, one clear suggestion and a reason.

Check that the size is in stock before you recommend it. Suggesting a large that's sold out forces a second round and leaves the customer feeling nobody looked. An AI agent connected to your Shopify catalog can cross-check the recommendation against real stock for each variant before replying.

When a person should answer instead

A person should answer when the question depends on judgment that isn't written down anywhere: bodies that fall between two very different sizes, tailored or formal pieces, and customers who already exchanged once for size and come back frustrated.

Hand off as well when the shopper asks for something the store can't promise, like "can you guarantee it fits?". The honest answer is that nobody can guarantee fit remotely, and that lands better in a teammate's voice, along with a clear explanation of the exchange policy.

A practical setup is copilot mode: the AI drafts the recommendation from the product data and a person approves or edits it before it goes out. It's useful in the first weeks, while your team confirms the fit notes are right, and on your most expensive product lines.

Common mistakes when helping shoppers pick a size

The most common mistake is answering with a size without checking the specific product, as if everything in the store fit the same way. Four more are worth avoiding.

Always sizing up "just in case". It lowers the risk of too small, but raises exchanges for too loose, especially in footwear.

Not explaining how exchanges work. Saying early that size exchanges are possible, and how, lowers anxiety and unblocks the purchase. If the customer still needs one, this guide on handling returns and exchanges on WhatsApp covers what comes next.

Not learning from exchanges. Every size exchange is data: if one style keeps getting swapped from medium to large, its fit note is wrong or missing. Reviewing those patterns monthly improves the product page and the chat answers at the same time.

Ignoring what other customers say. Reviews that mention fit ("runs small, size up") are the most credible source for the next buyer. Asking for photo reviews after delivery builds that archive naturally.

Frequently asked questions

What should I do if a customer is between two sizes? Ask how they like the fit and check whether the fabric stretches. If they prefer a relaxed fit or the fabric has no give, recommend the larger size; if they want it snug and the fabric stretches, the smaller one. Explain the reason in one sentence so the choice is theirs.

Isn't a size chart on the product page enough? It helps, but it doesn't replace the conversation. The chart works for people who know their measurements; many shoppers don't, or don't trust them, and would rather ask. The chart and your chat replies should draw on the same data.

Can AI recommend sizes on WhatsApp? Yes, as long as it has the right data: garment measurements, fit notes per style and stock per variant. Without that it can only repeat the chart. For borderline or high-value cases, have it hand off to a person or work in copilot mode.

◆ how heysteff does it

Steff answers size questions on WhatsApp using the fit notes in your knowledge base and your Shopify catalog, and hands off to your team when a case needs it.

Related

◆ next step

Answer every size question in minutes, with live stock in view.

Start free → Get a demo See pricing