AI Size & Fit Advisor: Personalized Sizing Recommendations
Automatically recommend the perfect size by asking customers qualifying questions and matching their measurements to your sizing charts.
Overview
Wrong size selection is the biggest driver of returns in eCommerce. Customers guess their size based on generic charts, order the wrong fit, and return it. You lose shipping costs both ways, processing time, and often the sale entirely when they don't reorder.
Zipchat's Size & Fit Advisor eliminates guessing by having conversations about fit. You upload your sizing charts as PDFs to the knowledge base and tag them to specific products. Then you prompt the AI on what questions to ask. For t-shirts: body type, height, fit preference (loose or fitted), gender. For shoes: current shoe size, reference brands they wear, foot width. The AI asks these questions naturally in conversation, matches the customer's answers to your sizing chart, and recommends the exact size they should order.
It also shares fit notes and exchange policies proactively. If a customer is between sizes, the AI explains which way to size based on their preferences. If your return policy is particularly customer-friendly, it mentions that to reduce purchase anxiety. The entire interaction happens in seconds, delivering confidence at the exact moment sizing questions would normally cause cart abandonment.
Brands using conversational sizing see 30-45% reduction in size-related returns within 60 days. Cart abandonment on product pages drops by 15-25% because customers get answers before doubt sets in. More importantly, customer satisfaction scores for fit increase significantly because people receive products that actually fit. For apparel and footwear brands, this single ability can save thousands in return shipping costs monthly while improving the customer experience and increasing first-order conversion rates.
Overview
Problem
Solution
Scenario description
Dress Shopping Uncertainty
A customer is looking at a midi dress and asks: "I'm between sizes. What size should I get?"
The AI responds: "I can help you find the perfect fit! Quick questions: What's your height, what size do you normally wear in dresses, and do you prefer a fitted or relaxed fit?"
Customer: "I'm 5'6", usually wear a medium, and I like it fitted but not tight."
AI checks the sizing chart and responds: "Based on your measurements and fit preference, I'd recommend size Medium. This dress runs true to size with some stretch in the waist. If you're usually between S and M, the Medium will give you that fitted-but-comfortable fit you're looking for. Our exchange policy is super easy if you need to swap sizes, but Medium should be perfect for you."
The customer orders with confidence instead of abandoning the cart. She receives the dress, it fits perfectly, and there's no return to process.
Sneaker Sizing Confusion
A customer messages about a new sneaker style: "Do these run big or small? I wear a 9 in Nike."
The AI asks: "Great question! Do you find your Nike 9s fit perfectly, or are they a bit roomy? And do you have narrow, regular, or wide feet?"
Customer: "My Nikes fit perfect and I have regular width feet."
AI: "Perfect! Our sneakers run about a half size larger than Nike. I'd recommend ordering size 8.5 for the best fit. If you prefer a little extra room, size 9 would work too, but 8.5 will match your Nike fit exactly."
The customer orders the right size immediately. No guessing, no ordering multiple sizes to try at home, no returns. Just accurate guidance based on real reference points the customer already knows.
Setup guide
- Upload your sizing charts as PDFs to Knowledge Base in your Zipchat dashboard
- Tag each sizing chart with specific product names or categories it applies to (e.g., "Women's Tops," "Running Shoes," "Denim Jeans")
- Navigate to AI Training → Prompt & Tools → Additional Instructions
- Write prompts explaining what questions the AI should ask for each product category (height, weight, body type, fit preference, reference brands, etc.)
- Include fit notes in your prompts (e.g., "this style runs small," "has stretch," "fitted through chest")
- Add your exchange/return policy details in the knowledge-base and add to the prompt something like "re-assure them citing our return policy", so the AI can mention them for reassurance
- Test with sample sizing questions across different product categories
Ability statistics
Technical table
FAQs
How does the AI match customer information to sizing charts?
You upload sizing charts as PDFs to the knowledge base and tag them to specific products. In your prompts, you tell the AI what questions to ask (height, weight, body type, fit preference, etc.). When a customer asks about sizing, the AI asks those qualifying questions, then cross-references their answers with your sizing chart data to recommend the specific size. It can handle complex scenarios like "between sizes" or "runs large/small" based on the fit notes you provide.
What information should I prompt the AI to ask customers?
It depends on your product category. For apparel: height, weight, body type, normal size worn, fit preference (fitted vs loose), gender. For shoes: current shoe size, reference brands they wear (Nike, Adidas, etc.), foot width, sock thickness preference. For accessories like rings or belts: specific measurements like finger circumference or waist size. The key is asking for information that customers already know rather than requiring them to measure themselves with a tape measure.
Can this handle different sizing charts for different products?
Yes. You upload multiple sizing charts and tag each one to specific products or categories. The AI automatically pulls the correct chart based on what product the customer is asking about. If you sell both men's and women's clothing with different sizing, or US vs EU sizing, or different fits (regular, slim, relaxed), each gets its own chart. The AI handles the logic of which chart to reference based on the conversation context.
Does this work across all channels like WhatsApp and Instagram?
Absolutely. Size & Fit Advisor works on chat bubble, WhatsApp, Instagram DMs, Messenger, and email. This is especially valuable on WhatsApp and social channels where customers are browsing from their phones and more likely to have sizing questions but less likely to dig through size charts on product pages. The conversational format works naturally in any messaging environment.
How much do sizing recommendations actually reduce returns?
Brands implementing conversational sizing see 30-45% reduction in size-related returns within 60 days. The financial impact is significant: if you're processing 100 size-related returns monthly at $15 average cost per return (shipping both ways, processing, restocking), that's $1,500 monthly. A 35% reduction saves $525 per month or $6,300 annually. Beyond direct costs, you also improve customer satisfaction, increase repeat purchase likelihood, and reduce the frustration that causes customers to never order from you again.
Try Zipchat in Action!
Enter your store URL to see how Zipchat would behave.



