How to build a Facebook Messenger Chatbot without coding
Building a chatbot can seem like a daunting task. This is especially true for those who (like me) lack coding knowledge. Thankfully, there are a number of free platforms such as Chatfuel, Microsoft Bot Framework, and my favorite, SnatchBot. Platforms like this let you get up and running with a great chatbot in no time. In this article, I’ll be walking you through the basics of setting up a Facebook Messenger chatbot in the SnatchBot platform.
Step 1: Establishing the purpose
Ideally, if you’re reading this article, you already have an idea for a chatbot you want to build. If not, try to think of some sort of utility that could be solved with a chatbot. In this case, I’ll be showing you how to create a chatbot for a local small business. This chatbot will be connected to Facebook Messenger, answering frequently asked questions and providing basic information. Let’s call it “Joe’s Gym”.
Step 2: Connecting the chatbot
In order for a chatbot to function properly, it must be connected to the channel you want to deploy it on (Messenger in this case). How this is done varies based on both the platform where you build the chatbot, as well as the channel where it is executed. SnatchBot provides an excellent overview of how to do this for Facebook users. It’s straightforward, just be sure you are complying with Messenger’s platform policies here. Once you click on the Facebook option, you’ll see your existing business Facebook pages to choose from, or you can create a new one.
Step 3: Building the chatbot
This is where the fun starts. First, let’s make sure we identify what the chatbot will be capable of. Here’s a short list of what I’ll be including:
Ability to answer/provide:
- Phone number
- Location
- Business hours
This is fairly straightforward, and provides a basic understanding of how to get up and running.
The first thing I’ll do is create my greeting message. We’ll want something simple and friendly. We also want it to encourage some sort of user action. Let’s go with “Hi there, this is Joe’s Gym. How can I help you today?” This is done by creating a Bot Statement interaction within the chatbot.
Important to notice: The checkmark beside “start interaction” causes this message to be the first to display when a user enters the conversation.
Next, we’ll build the individual content interactions that contain the information users are looking for. As stated above, we’ll create outputs for hours, location, and contact info. We’ll also want a goodbye message. For each of these, we’ll create a Bot Statement Interaction.
Hours
Location
Contact
Goodbye
Step 4: Connecting the conversation
We now have all of our chatbot’s content prepared. However, to make it functional, we must first connect the interactions. This consists of creating if/then statements that tell the chatbot which user inputs should trigger specific interaction content (location, hours, etc.).
To ensure that users move through the conversation smoothly, we’ll be using Quick Replies. These are words that you suggest the user inputs to gather the info they are looking for. For example, our greeting message (pictured below) suggests three options: Hours, Location, and Contact.
Quick Replies can be selected and entered (below) at the bottom of the Edit Interaction module.
You’ll notice that a blue box is ticked in this image. This prevents users from entering their own dialogue, therefore streamlining the flow and preventing them from inputting info the chatbot doesn’t recognize (which could cause it to fail).
Once your Quick Replies have been entered, scroll up the dialogue box to add a connection. Each connection you add should point the name of the Quick Reply you just entered to the corresponding interaction. Below is an image of how this is done for Hours.
Complete this for each interaction to finish setting up your chatbot.
Step 5: Testing connections
The best practice is to test your chatbot as you build it to identify and eliminate flaws. SnatchBot let’s you do this within the platform (pictured below).
In addition to this, it’s also a good idea to test the chatbot in the channel where it will live. In this case, I repeat the same tests but in the Facebook Messenger platform.
Conclusion
That’s how easy it is! If you’d like to test the chatbot we just built, you can do so here.
This tutorial should let you build your first chatbot on the SnatchBot platform and have it up and running in 30 minutes. Best of luck, and happy building!
