E-commerce
June 28, 2026
A size recommendation can reassure the customer and reduce returns, but it must be explained. The customer wants to know what the recommendation is based on: measurements, history, size guide, reviews, fit, or data from a previous purchase.
The chatbot must use this information with transparency and caution. It recommends a size, explains the reason, points out the limitations, and respects privacy preferences.
This guide shows how to use an AI chatbot to recommend sizes without giving an impression of magic or surveillance.
Summary
Why explain the recommendation?
A recommended size without an explanation can be useful, but it can also seem arbitrary. If the customer hesitates, they need to understand why the chatbot is advising S, M, or 42.
The chatbot must make the recommendation legible: data used, product fit, comfort preference, and any potential margin of uncertainty.
A size recommendation inspires confidence when it explains its reasoning in simple language.
Convert over 2,000 customers on average per month with Qstomy.
The world’s 1st Shopify AI dedicated to customer conversion
Empowering 200+ e-commerce merchants
Which data should be used?
The bot can use the size guide, the measurements provided, the usual size, feedback from other customers, purchase history, product fit, and material.
It must limit the use of history to useful elements and explain if a personal data point influences the recommendation.
How do you manage customer history?
If the customer has already purchased a similar size, the chatbot can use it to improve the recommendation. It must state it simply: "I am also basing this on your previous purchase in this category."
If the customer does not wish to use their history, the bot must offer a recommendation based on measurements or the standard guide.
How should the recommendation be presented?
The answer must give a main size and, if necessary, an alternative according to the desired fit. For example: "Take M for a slim fit, L if you prefer more room."
The chatbot must avoid presenting a size as certain if the customer is between two measurements.
How to manage limits?
Sizes vary by brand, cut, and material. The bot must remind the user of the exchange conditions, especially if the product is on final sale, personalized, or difficult to return.
The recommendation should help with the choice, not hide the risk of a non-exchangeable product.
Which flow to follow?
The flow must recommend transparently.
Identify product, category, guide, fit, material, and hesitated size.
Collect measurements, comfort preference, and useful history if authorized.
Compare data with the guide and available product feedback.
Provide an explained recommendation with an alternative if necessary.
Escalate inconsistent measurements, non-exchangeable products, volume orders, and expert requests.
Which messages should be used?
To recommend: “With your measurements and the cut of this model, size M seems the most suitable.”
For history: “I can use your previous purchase to refine the recommendation, if you wish.”
For limit: “Since you are between two sizes, the choice mainly depends on the desired fit.”
When to transfer?
Transfer is necessary if measurements are inconsistent, if the product is not exchangeable, if the customer requests highly personalized advice, or if a professional order depends on sizing.
The bot must transmit the product, measurements, whether history was used or not, recommendation, alternative, and remaining doubt.
Which KPIs should be monitored?
Follow accepted recommendations, returns for sizing, avoided exchanges, refusal to use purchase history, sizes between two measurements, and post-reception satisfaction.
This data helps improve the size guides and ensures that the recommendation does not create false confidence.
Which mistakes should be avoided?
Avoid using history without transparency, guaranteeing the size, ignoring the cut, or recommending without recalling the limits of exchange.
The chatbot must advise with precision and humility.
How can Qstomy help?
Qstomy can connect the chatbot to the catalog, variants, size guides, stock levels, SMS campaigns, social content, and support rules to answer clearly, then hand over sensitive cases with an actionable summary.
The chatbot helps the customer move forward without inventing a reference, size, availability, or promotion that has yet to be confirmed by a reliable source.
Explore AI support, the AI sales agent, or request a demo.
Key takeaways
Takeaways
A size recommendation must explain the data used, the fit, the material, and the level of uncertainty.
What the customer needs to understand
The customer must understand why a size is being suggested and how to choose if they are hesitating between two options.
The chatbot's correct limit
The chatbot can recommend, but it must transfer expert cases, inconsistent measurements, and non-exchangeable products.
Enzo
June 28, 2026
