How I picked the best AI chatbots for Australian businesses
Every chatbot here got judged on the same five things, in this order:
- Does it answer from your own information? Your products, prices, stock, hours and policies. A chatbot that only knows the internet is a liability.
- Does it resolve or just collect? The whole point is answering the question at 9pm, not taking a number for Monday.
- How much work is setup? Most small business owners don't have a spare week to train a bot.
- What does it really cost? Flat monthly fees beat per-conversation or per-resolution billing, which can blow out on a busy month. Prices are as each company lists them.
- Does it suit an Aussie business? Australian dollars, support in our time zone and a sensible approach to customer data all count.
What is the best AI chatbot for small business in Australia?
Matey is the best AI chatbot for most small businesses in Australia, because it's built for you on your own products and policies, it books and sells after hours, and it costs a flat AU$149 a month after a free first month. If you'd rather build a chatbot yourself, the runners-up below are all solid, and each one wins for a particular kind of business.
Which AI chatbot should you pick for your business?
It comes down to how much time you have and what the chatbot needs to do:
| If you... | Pick |
|---|
| You want it built for you, on your own products and prices | Matey |
| You want a free chatbot and don't mind setting it up | Tidio |
| You run a support team with a help desk | Intercom Fin |
| Your business already runs on HubSpot | HubSpot |
| You want a chatbot trained on your site in under an hour | Chatbase |
| Most of your customers message you on Instagram | ManyChat |
| You're technical and want something fully custom | Botpress |
Wrapping up
A website chatbot has one job: answer the customer before they go somewhere else. If you want that done properly without spending your evenings training a bot, start with Matey. The first month is free, so the only thing you risk is finding out how many customers were asking questions at 9pm.
If you're a hands-on type who'd rather build it yourself, the runners-up above are all good tools. And if you're curious how these things understand questions in the first place, my guide to natural language processing explains it in plain English.