Look at most Australian retail websites and you'll find the AI assistant in the same spot: a chat bubble in the bottom corner, or a link on the contact page. It answers questions about delivery and returns and deflects some support tickets, and it does that job reasonably well.
But the shoppers it talks to have mostly made up their minds already. Either they've bought and want to know where their parcel is, or they've given up looking for an answer and want a person.
Whether a sale happens is decided earlier, and most retail chatbots aren't there when it does.
Sales are lost at the point of decision
Baymard Institute puts the average online cart abandonment rate at just over 70%, based on 50 separate studies. Some of that can't be avoided, because plenty of shoppers are only browsing.
A lot of it can. When Baymard asked shoppers why they left during checkout, 40% said the extra costs were too high, 20% said delivery was too slow, 13% didn't like the returns policy and 12% couldn't see the total cost up front.
Many of those reasons come down to a question the shopper couldn't get answered when it mattered. How much is shipping to my postcode? Will it arrive before the weekend? Can I send it back if it doesn't fit? A clear answer at that moment keeps a good share of those shoppers in the checkout.
Physical stores have the same problem in a different form. Zebra Technologies' 2025 Global Shopper Study found that 52% of shoppers had recently left a store without everything they came for, often because items were out of stock or hard to find.
In a store, an unanswered question usually ends with "I'll think about it" and a walk to the exit.
A chatbot on the contact page meets none of these shoppers. By the time someone goes looking for "Contact us", many have already decided to leave.
Two places retail AI earns its keep
Product and category pages
This is where shoppers compare options and talk themselves into or out of a purchase.
The questions are specific: does this run small, and is this jacket waterproof or only water-resistant? An agent trained on your product catalogue can answer in seconds and suggest a better fit or a matching item. That is where conversion and basket size move.
Checkout
Hesitation at checkout is expensive because the shopper has already done the hard work of choosing.
Their questions are about delivery, payment, returns and the final cost.
An agent that can answer "will this arrive by Friday?" with real delivery information, or explain the returns policy in plain language, removes the last reason to leave.
What we've seen in retail
Our clients see an average 25% uplift in conversion rate from AI sales and support agents. The biggest gains come when the agent is placed where shoppers are making decisions, alongside the service enquiries it already handles.
Co-founder of TWL, puts it this way: "Our AI agent drives sales 24 hours a day, 7 days a week. Our customers love it because they can get the information they need, when they need it."
Find your hesitation points first
Before choosing any technology, work out where your shoppers hesitate today. Most retailers already have the data:
- Checkout analytics: Which step loses the most people? Does it line up with delivery, payment or returns information?
- Support enquiries: Which pre-purchase questions arrive most often by email, chat and phone? Each one is a question the product page didn't answer.
- Store team feedback: What do floor staff get asked most? Which of those questions make them walk away to check?
- Site search logs: What are people searching for and not finding?
Pick the two or three moments with the most volume and the clearest link to a lost sale, and put an AI agent there first. Your contact-page chatbot can stay, but it shouldn't be the only place a shopper can get an answer.
Where to start
If you'd like help finding those moments in your business, our AI Advantage Strategy is a short, expert-led session with your team.
Together we'll map where your highest-ROI opportunities sit and what it would take to act on them.
Sources
- Baymard Institute, Cart Abandonment Rate Statistics 2026
- Zebra Technologies, 18th Annual Global Shopper Study (2025), as reported by MarketScreener
