Can the chatbot answer questions about product sizes and colours?

Yes. You load your product details, sizes, and colour options into the knowledge base, and the bot pulls that information when a customer asks. The more complete your product info, the better the answers.

What happens when the bot cannot answer a question?

The bot routes the conversation to your staff. You set the conditions, such as when a customer asks about a custom order or a payment issue, and the bot hands it over cleanly.

Does the bot show live stock levels from my POS system?

Not automatically. Stock answers come from what you load into the knowledge base. If you update stock info weekly, the bot reflects that. It does not connect to a live POS in real time.

Which channels can the bot reply on?

WhatsApp, Instagram, Telegram, email, and your website chat, all from one inbox. Your customers message wherever they prefer, and the bot replies on the same channel.

Does Polaris work for both physical stores and online shops?

Yes. Physical stores use it for store hours, location, and in-store stock questions. Online shops use it for shipping basics, product details, and order inquiries, before routing buyers to staff for payment.

February 14, 2026 • Jiun

How a Retail Chatbot Answers Your Customers in Seconds

A retail chatbot answers your customers’ product, stock, size, price, and hours questions in under 5 seconds, any time of day. For a busy fashion boutique in KL or a gadget shop in Penang, that means fewer lost sales and less time spent replying to the same messages over and over.

Here is why that matters, and how it works in practice.

Why slow replies cost Malaysian retail shops real money

A customer messages your Instagram asking if a dress comes in size XS or whether you have a specific phone model in stock. If you do not reply within minutes, they move on. They find another shop that answers faster, and that sale is gone.

Research on lead response time shows that leads contacted within 5 minutes are 21 times more likely to qualify than those contacted after 30 minutes. That number comes from a 2011 Harvard Business Review study on online sales leads, and the pattern holds true for retail messages today. The gap between a fast reply and a slow one is not small.

For most shop owners, the problem is capacity. You are folding clothes, attending to a customer at the counter, or packing an order. You cannot watch your phone at the same time. Messages pile up. By evening, some of those buyers are already unboxing something they bought from a competitor.

What a retail chatbot actually does

A retail chatbot is a bot that replies to customer messages automatically, day and night, from your shop’s own information. A customer asks “do you have this in navy?” or “what time do you close on Saturday?” and the bot answers from what you have loaded into its knowledge base.

For retail, the knowledge base typically covers:

Product names, descriptions, sizes, and colour options
Price ranges or exact prices
Stock status (in-stock, low stock, out of stock, based on what you have loaded)
Store hours, location, and parking info
Shipping basics for online orders
Return or exchange policies

The bot does not invent answers. It searches the knowledge base you built, pulls the relevant detail, and replies in plain language. If the question falls outside what you have loaded, it tells the customer it will connect them with staff.

A real scenario: a fashion boutique in Petaling Street

Take a women’s fashion boutique in Petaling Street, KL. They carry 80 to 100 SKUs at any time, with multiple sizes and colours per item. Their WhatsApp gets around 60 to 80 messages a day. Most are variations of the same questions: “do you have size S?”, “what is the price for this dress?”, “are you open on Sunday?”.

Before the bot, the owner or her single staff member answered every one of those manually. Evenings were the worst, after the physical store closed but before the online crowd stopped browsing. Replies came 12 to 18 hours later. Some customers had already bought elsewhere.

After loading their product catalogue, hours, and location into the knowledge base, the bot handles the routine questions instantly. A customer messages at 11pm asking about a blouse in dusty pink, size M. The bot checks the knowledge base, confirms it is available, and sends the price. The customer says she wants to order. The bot captures her name and contact number, then flags the conversation to the owner to confirm and arrange payment in the morning.

The sale was captured. The owner slept.

How the bot knows when to hand off to staff

Not every message is a product question. Some customers want to negotiate, ask about custom alterations, or have a complaint. The bot is not the right tool for those.

You configure hand-off rules. When a buyer says they are ready to pay, the bot routes them to staff. When a question falls outside the knowledge base, the bot routes it. When a customer explicitly asks to speak to a person, the bot routes it.

The staff member picks up the conversation in the same inbox, with the full message history visible. There is no need to ask the customer to start over.

Research from Meta and Kantar in 2025 found that 73% of consumers prefer messaging a business over calling or emailing, and 72% are more likely to buy from brands that offer messaging. The bar for being a responsive business is already set. The question is whether your shop meets it.

What the bot does not do

A few things worth knowing before you set one up.

The bot does not process payments in chat. It can capture buyer details and flag a ready buyer to staff, but the payment step happens outside the bot, on your payment link or at the counter.

Stock accuracy depends on your knowledge base. If you update stock info once a week, the bot reflects week-old stock data. It does not connect to your POS or inventory system in real time. Some shops update the knowledge base daily; others do it when a product sells out. How current your stock answers are depends on how often you update.

The bot also does not handle complex negotiations or complaints well. Set your hand-off rules clearly so those conversations go to a person fast.

Setting it up for a retail shop

The setup comes down to three things: building the knowledge base, connecting your channels, and setting your hand-off rules.

For the knowledge base, quality matters more than volume. A clear, well-written product list with accurate sizes, colours, prices, and stock notes gives better answers than a messy spreadsheet with half-complete entries. Take time with it.

For channels, most Malaysian retail shops start with WhatsApp because that is where most customer messages already come from. Adding Instagram and Telegram is straightforward from the same inbox. You do not need a separate tool for each channel.

For hand-off rules, think about the moments in a conversation where a human adds real value. Set those as your triggers.

If you want to compare how AI chatbots work for e-commerce versus physical retail in Malaysia, that post covers the channel differences in detail. For a broader overview of what to look for in a Malaysian chatbot setup, the AI chatbot guide for Malaysian businesses is a good starting point. And if you are thinking about pricing, the WhatsApp chatbot pricing breakdown in RM walks through what different plans cover.

Polaris is billed in RM, connects WhatsApp, Instagram, Telegram, email, and your website into one inbox, and replies in under 5 seconds around the clock.