Small Business Tips
Over the last few years, messaging and chat applications such as WhatsApp and Snapchat have come on the scene. Suddenly, chat is the new email.
Millennials now use chat apps rather than email to communicate. And no wonder: chat works seamlessly from mobiles and it’s packed with helpful features.
In business, chatbots for websites are the virtual assistants that make communications with customers easier. Let’s find out more.
What’s a chatbot?
So, what exactly is a chatbot?
Let’s pretend you’ve just landed your dream promotion. Well done. But now you need to change the information on your business cards.
To do this, you start a messaging conversation with the chatbot on your firm’s messenger app. You find that, unlike the robotic, command-centered robots of the past, your chatbot is super-friendly. He or she uses words like ‘sure’ and ‘no problem.’ It’s like talking to another human.
Remarkably, your chatbot can pull up the template from your last order. He knows you. He notices that your job title has changed.
Pretty smart. From a user perspective, you’ve been able to order your new business cards quickly, without having to fill out any forms, log onto any specific applications or learn how to use any new apps. Best of all, you’ve been able to use natural language to order your new cards.
Chatbot e-commerce stats
No surprise, then, that chatbots could well be the future of e-commerce and marketing. Facebook recently released data that illustrates the value of chatbots for business websites:
- People and businesses exchange two billion messages each month
- 56 percent of people prefer to message than speak to customer service
- More than half of people are more likely to buy from an organization they can message
Want some more stats? Business Insider experts believe that 80 percent of enterprises will use chatbots by 2020. And research author Lauren Foye says that banks will be able to use chatbots to automate up to 90 percent of their customer interactions by 2022.
A recent Spiceworks survey suggested that 40 percent of companies with over 500 staff planned to bring in a chat robot for their corporate mobile devices this year.
The latest figures from Drift tell us that 27 percent of adults are ready to use a chatbot to buy basic goods. And 13 percent have bought high-ticket items using chatbots. Adobe informs us that nearly a third of leading businesses use AI for marketing and another 31 percent plan to do so in the next year.
Opus Research predicts that, by 2021, $4.5 billion will be invested in chatbots.
What are chatbots used for?
Drift’s report, The State of Chatbots, cites a survey that predicts the most common uses of chatbots will be as follows:
- Getting a quick answer in an emergency (37 percent)
- Resolving complaints or problems (35 percent)
- Getting detailed answers or explanations (35 percent)
- Finding a human customer service assistant (34 percent)
- Making a reservation (33 percent)
- Paying a bill (29 percent)
- Buying a basic item (27 percent)
- Getting ideas and inspiration for purchases (22 percent)
- Adding yourself to a mailing list (22 percent)
The standout results here are that customers want chatbots to find quick answers to questions, resolve issues, provide detailed answers for questions and help them find a human customer service agent.
How do chatbots work?
Obviously, we could get very technical here — if we understood artificial intelligence that well. But we don’t. So let’s keep it simple.
Let’s say you’re on Facebook’s Messenger chatbot. You find a product that you want to buy — maybe some beans. You send the chatbot a message, which gets picked up by the backend.
The AI-driven chatbot uses natural language processing (NLP — this is how computers read language) to convert your plain text into code.
The goal of the chatbot is to get out of this conversation as quickly as possible by closing down all the data. It sends your information into a decision engine, as it needs to know the quantity of beans you’re after.
Your e-commerce chatbot then uses natural language generation (which is how computers write) to ask you how many cans of beans you want. Your reply goes back through NLP into the decision engine.
Here’s the clever part. The chatbot will mine its pre-existing data about the beans you need and which stores stock them. It’ll then give you the price.
You’ll then be directed to a payment portal. After getting the go-ahead from the gateway, the chatbot will place your order. A couple of days later, your beans will arrive.
