Lesson 254 · 2 min

Replying to customers in chat

AI can help you build a set of ready replies for the questions customers ask every day, so you answer faster and still sound kind.

Look at your Zalo, Messenger or Shopee chat and you will see the same questions again and again: price, size, shipping fee, how long delivery takes, can I return it. Typing each answer fresh eats hours.

Copy ten common questions, remove names and phone numbers, and ask AI to write short, warm replies in your voice. Save the good ones as quick replies in your chat app, most of which have this feature.

Keep a person in charge. Use the saved replies as a starting point and add the customer's name or a detail from their message. For anything unusual, like a complaint or a big order, answer yourself.

Be careful with automatic chatbots that answer for you. They can promise a discount you never offered or give the wrong delivery date. If you try one, check its answers every day at first.

Do it now

Write your first ten quick replies

  1. 1. Scroll through last week's chats and copy the ten questions you see most.
  2. 2. Remove every name, phone number and address.
  3. 3. Paste them into a chatbot with the prompt below, then edit the replies until they sound like you.
  4. 4. Save them as quick replies in your chat app.
I run a small shop selling [WHAT YOU SELL] in [CITY]. Here are the questions customers ask me most often. Write a short, friendly reply for each one (under 40 words), in a polite, warm tone. Use only these shop facts and leave a [BLANK] where you are not sure: shipping fee [AMOUNT], delivery time [DAYS], return policy [POLICY], payment [BANK TRANSFER / COD]. Questions: [PASTE QUESTIONS].

Quick check

Before pasting real customer chats into a chatbot, what should you do first?

Before pasting real customer chats into a chatbot, what should you do first?

More on this

Words from this lesson

Data privacy
Keeping personal or confidential information under control, including deciding what you type into AI tools and where it is stored.
Human in the loop
Designing a process so a person reviews or approves the AI's work before it has real consequences.