Using chat robots is one of the hottest areas of Internet marketing these days. A chat robot (chatbot) is a piece of software that is installed on a messaging platform and interacts with your visitors automatically. They often provide 24×7 unattended customer service. They can be fun to talk to as well!

Chatbots are often deployed in the form of chat window on the website but are now increasingly installed in other messaging platforms, like Facebook Messenger, Apple iMessage, Twitter, or Slack. Some chatbots use a voice interface instead of a text interface, which is the case with Amazon Alexa, Apple Siri or Microsoft Cortana.

Chatbots are often based on artificial intelligence (machine learning) and can understand the meaning of written text. When someone interacts with your chatbot, it responds with intelligent answers.

Some chatbots are more conversational than others. For B2B usage, it is recommended your bot is more focused on providing concise answers than being very conversational.

It is important to understand that the chatbot is not a service in itself; it is merely an interface to the service, much like a website that offers flight booking services. Thus, you need to train the chatbot to provide some meaningful answers.

Many chatbot development platforms exist to build chatbots, and they are all targeted at different combinations of messaging platforms. Some chatbot development tools are:

• ChattyPeople
• Chatfuel
• FlowXO
• BotKit
• Octane.ai
• Motion.ai
• MEOKAY
• Smooch
• Botsify
• Beep Boop

For example, I built a chatbot using Motion.ai that I installed on one of my websites, although this chatbot could be deployed on Facebook Messenger or other platforms as well. This is what it looks like om the website:

Chatbot

Chatbot development platforms provides a generic chat robot that understands written text. But to make it meaningful to your business, you need to train it with company or domain specific knowledge.

This can be done by teaching it “hard coded” knowledge (i.e. static knowledge fed into the robot at the time of training it), or by integrating the chatbot with external IT systems that can provide live dynamic knowledge.

The latter can, for example, be used to ask a robot the estimated time of arrival of a particular airborne flight, or the status of a support ticket. Answering questions like “What is the price of <productname>?” can be either static hardcoded knowledge, or retrieved dynamically using integration to external systems using webhooks.

It is obviously easier to get started with a chatbot that is only trained with static knowledge without live integrations to external database systems.

You may be thinking it is incredibly expensive and complicated to build a chatbot based on artificial intelligence. It isn’t. The screenshot below shows how Motion.ai can be used to develop a chatbot:

Chatbot development

Some chatbot development platforms offer free tiers you can play with. Unless you have a lot of people interacting with your bot, the recurring cost may not be more than 50 dollars a month, or even less. The price usually depends on the monthly number of audience interactions you have.

Some work is needed to set up the robot in the first place of course, and to improve or change its behavior over time.

If you want to provide automated 24×7 customer service on your website or a messaging platform like Facebook Messenger or Slack, a chatbot may be worth looking into.

What knowledge should you train it with? My suggestion is to browse your previous email or support ticket history and teach the robot to answer the most common questions you have received in the past. At least, that is a good start!

Do you want me to develop a chatbot for your business? Contact me!

Facebook Comments

Learn how to build an online business!

Monetize_knowledge_vertical_banner

Register now to get the FREE eBook "Turn your knowledge and skills into a profitable online business" and our latest content by email.

Powered by ConvertKit
Share the love!Email this to someoneShare on FacebookTweet about this on Twitter