ChatonboChatonbo
Solutions
TemplatesTry ItUse CasesPricingCompareIntegrationsDocsBlog
LoginSign Up Free
BlogResearch
ResearchSeptember 19, 20266 min read

Order Tracking Questions in Store Chat: 4.2% of Messages, 16% Answered Badly (2026 Data)

EK
Elena Kowalski
Head of Customer Success
Order Tracking Questions in Store Chat: 4.2% of Messages, 16% Answered Badly (2026 Data)

On this page

The numbersWhy it failsWhat a good answer needsThe bigger pictureFAQ

TL;DR: "Where is my order?" is 4.2% of everything shoppers type into a live store chat, the fourth most common intent, and the one assistants get wrong most often: 16% of order-status questions in our 60-day sample ended without a usable answer, and those conversations collected the most thumbs-down. The fix is not a smarter model. It is a real connection to the store, asking for the order number and the email in one message instead of two, and an answer that carries the carrier link.

The numbers

Two data sets, both from Chatonbo production:

  • 9,480 visitor messages on live storefront assistants, 60 days to September 19, 2026: order status was 4.2% of messages. Among the transactional intents only discounts (7.0%), product recommendations (4.7%) and price (4.6%) were more common.
  • 5,470 conversations across 246 assistants of every kind, 90 days to September 11, 2026: order status was 2.0% of conversations overall and 7.8% on Shopify stores.

The failure rate matters more than the share. Order status failed 16% of the time, against 11% for discounts. And 2.5% of order-status answers received a thumbs-down, the highest of any intent.

Why it fails

We read the failed conversations. Three patterns cover nearly all of them:

  1. No connection to the store. The assistant only knows the pages it crawled, so the best it can do is quote the shipping policy. A shopper who asked "where is my order" and got "orders usually ship within 2 business days" counted as a failure, and they rated it that way.
  2. Two questions instead of one. The assistant asks for the order number. The shopper leaves. Or the shopper gives the number and the assistant then asks for the email. Every extra turn loses people. On Business-plan stores in our sample, 105 of 286 order questions were met with a request for details.
  3. Sloppy matching. The shopper types "order #1042" or "no. 1042" and the lookup searches for the literal text. Or the number is mistyped and the email, which was correct, is never used.

What a good answer needs

  • A live lookup against the store, not a policy quote. On Shopify that is the Chatonbo app; on WooCommerce and Wix, the plugin and the app.
  • One message that asks for the order number and the checkout email together, with either one enough to find the order.
  • Number normalisation ("order #1042" is 1042) and a retry by email when the number does not match.
  • The carrier name and a tracking link in the answer, plus the estimated delivery date when the store recorded one. When the store attached a number but no link, the assistant should build the carrier page from the number.
  • A privacy rule: the item list and the tracking link only when the shopper gave both the order number and the email of that order. With one of them, the assistant gives status, carrier and dates and asks for the other detail. Sequential order numbers and public email addresses should not unlock someone else's order.

All of that is how Chatonbo assistants answer since September 19, 2026, and it is the reason the failure rate is the number we watch each week.

The bigger picture

Order tracking is table stakes in 2026: every serious store assistant claims it. The difference is in the failure rate, and you only see the failure rate if you measure it. The figures above, with their data sets and dates, are kept current on AI chatbot statistics 2026; the wider picture of what shoppers ask is in the full study.

FAQ

How often do shoppers ask a chatbot where their order is?

On live Chatonbo storefront assistants, 4.2% of 9,480 visitor messages over 60 days asked about order status, tracking or delivery. In the wider 90-day study of 5,470 conversations across all site types, order status was 2.0% of conversations overall and 7.8% on Shopify stores.

Why do chatbots fail at order tracking?

Three reasons in the data: the assistant has no connection to the store and can only guess; it asks for the order number and the email in two separate turns and the shopper leaves after the first; or the shopper types "order #1042" and the lookup does not normalise the number. In our sample 16% of order-status questions ended without a usable answer.

What does a good order tracking answer include?

The current status, the carrier and a tracking link, the estimated delivery date when the store has one, and the item list. For privacy, the item list and the tracking link should only be shown when the shopper gives both the order number and the checkout email; with one of them, status and carrier are enough.

EK

Written by

Elena Kowalski

Head of Customer Success · Chatonbo

Customer Success lead. Turns chatbot deployments into measurable business outcomes.

No signup required

See it work on your own website

Paste your URL and chat with an AI agent trained on your content - right now, in 60 seconds.

Try It on Your Website

Ready to try it yourself?

Deploy an AI chatbot on your website in 2-3 minutes.

Get Started for Free

Related articles

Do Shoppers Ask Chatbots for Discounts? Yes, More Than Anything Else (2026 Data)
September 19, 2026

Do Shoppers Ask Chatbots for Discounts? Yes, More Than Anything Else (2026 Data)

On live online stores, 7.0% of all visitor messages ask for a discount or promo code, ahead of product recommendations (4.7%), price (4.6%) and order status (4.2%). Here is what the data says and how a store should answer.

Do Website Visitors Want a Human Instead of a Chatbot? Only 1% Ask (2026 Data)
September 19, 2026

Do Website Visitors Want a Human Instead of a Chatbot? Only 1% Ask (2026 Data)

The fear that visitors reject an AI assistant does not show up in real conversations: 1.0% of 5,470 chats asked for a human, and 1.3% of storefront messages. What matters is what happens to that 1%: before we changed it, 40% of those requests dead-ended.

After-Hours Website Chat Statistics 2026: 39% of Conversations Start When Nobody Is Working
September 11, 2026

After-Hours Website Chat Statistics 2026: 39% of Conversations Start When Nobody Is Working

We analyzed 5,470 real website chat conversations. 39% started outside 9:00 to 18:00 local time and 24.7% on a weekend. Here is what those visitors ask and what a small team should do about it.

ChatonboChatonbo

AI-powered chatbot platform for sales, support, and lead capture.

Rated 5.0 out of 5 on the Shopify App Store (4 reviews)

© 2026 Chatonbo. All rights reserved.

Product

  • How It Works
  • Pricing
  • Lead CRM
  • Integrations
  • Documentation
  • Templates
  • Compare
  • Alternatives
  • Head-to-head comparisons
  • Research
  • Chatbot statistics
  • Customers
  • AI Visibility Checker
  • Blog

Solutions

  • Sales Agent
  • Support Agent
  • Shopify Agent
  • WooCommerce Agent
  • Wix Agent
  • Webflow Agent
  • BigCommerce Agent
  • Squarespace Agent
  • E-commerce Agent
  • Lead Generation Agent
  • All Industries
  • Small Business
  • Lead Capture
  • Real Estate
  • Healthcare
  • Restaurants
  • Legal
  • Education

Guides

  • Zapier Integration
  • Webhook Setup
  • WordPress Install
  • Shopify Install
  • Wix Install
  • HubSpot CRM
  • Slack Notifications
  • Zendesk Escalation
  • Zoho CRM
  • Google Analytics

Company

  • About
  • Contact
  • Help Center
  • Privacy Policy
  • Terms of Service
  • Affiliate Program
  • Chatbot Development Services
  • Authors
LanguagesEnglishEspañolPortuguêsFrançaisDeutschעברית