Building a chatbot on your own is not complicated, but it does require planning and anticipating the information that your chat users are looking for. Here are the main 10 steps to building your chatbot.

Step 1 – What is the purpose of your chatbot

Chatbots have several purposes such as:

  • Lead generation
  • Sales, online and offline
  • Improving usability and navigation of the site
  • Gaining insights
  • Support

Depending on the purpose of your chatbot, you need to engage the user. Understanding how well the chatbot is working is determined by the matrices you watch. For more information on how to measure the success of your chatbot, check out our blog “How to Measure Your AI Chatbot Success”.

Step 2 – Create a bubble diagram of conversations

The architecture of the chatbot has many tangents and sub-topics just like typical conversations. A quick way to organize this is to create a bubble diagram showing different conversations and if they are connected.

Another option to organize your thoughts is a tree diagram. Seeing the possible different paths that users can take is helpful to funneling them down to complete a purchase or filling out a lead form.

Step 3 – Create your default introduction question

The default introduction question needs to engage the website visitor. The more time that the chat user engages with your bot will increase the probability that they will purchase or provide contact information. This question directly affects the conversion rate and effectiveness of the chatbot.
Here is an example if you had a car rental company:

Option one:

Hi, I am a chatbot. What are you interested in? Select an option below:

  • Cars
  • Trucks
  • Family Vehicles
  • Insurance Information
  • Policies

This makes the user have to choose an option based on what they are looking for. They are then asked more clarifying questions to narrow down exactly what they need.

Option two:

Hi, I am a chatbot. How can I help you? Type in a question or what you are looking for.

With this option, the user has the freedom to type in an answer so that they get a specific question answered. The chatbot listens for a keyword like truck, car, van, etc. to determine the correct response. It provides the customer with exactly what they need without spending time trying to determine what they need solely through button clicks.

When the chatbot does not pick up any keywords in the user’s response, the chatbot “hits the rails” which is explained further, below. Additionally, this provides the most insight on your site because they can tell the bot about what obstacles they are facing during their shopping journey.

Option three:

Hi, I am a chatbot. What are you interested in? Type or select an option below:

  • Cars
  • TrucksFamily Vehicles
  • Insurance Information
  • Policies

The user has the choice to select one of the options that are given or they can ask the bot a question. This provides simple, quick information. This is the best option to engage the user because it gives them the option to explore what the company has to offer along with answering any questions at any point along their shopping journey.

After the default question, there are smaller branches that can be set up similarly to the default. Some examples of this are:

What kind of terrain are you planning on driving through?

How many people are traveling in this vehicle?

Chatbot Structure

This is a good time to take a tangent and explain how chatbots work.

Chatbots are trying to understand the “intent” of what the user is talking about.

Intents have two components:

Listeners – Listeners are keywords and phrases that trigger when that particular term is detected. Many listeners are already preprogrammed and are constantly being improved so you only have to think of the listeners that are relevant to your business. The keywords can use regex (a regular expression) logic such as:

word|words – list of words

^word – must begin with this word

\b word \b – must contain this phrase

A regular expression (regex) is a special text string to describe a search pattern. A regular expression is like an amplified wildcard.

Replies – Replies are the action that happens when the intent is triggered. There are a number of reply options that can be combined.

Connect to a question

Provide a text reply

Provide a button with a link

Show an image or video

Intents can be global or specific to a question. If the chatbot does not understand a question, an intent is triggered to reply with your fallback question as a last resort. There are other logic options such as “stay within this conversation.”

Step 4 – Outline your questions and conversational logic

After the default question, there is typically additional conversational logic. The conversational logic is like a tree where they branch into many different paths that the user can journey through. For example, if the default question was asking about types of vehicles there could be trucks, cars, bikes, etc. From there, there can be questions about the size, amount of miles per gallon, price ranges, etc. This takes the uses through a funnel to get them to exactly what product fits their request the best.

Outlining your conversations using a diagram can help with understanding what replies are going to which questions and other replies. Remember each of these conversions can have its own unique intents.

Step 5 – List your intents with common questions and replies

The global intents are part of the main paths that the user can take. These are the smaller branches of a tree to give them the most relevant information or products for their situation. When adding your frequently asked questions, enter them into the assistant intents.

The assistant intents are made so that the replies will be brought up if the bot does not trigger any global intents. This is where the frequently asked questions will be stored.

For example, when a chat user is trying to find the phone number of the support line the bot will pick up the listener and the appropriate reply. Typically, when the chat user is on an assistant reply, the back button goes to the default question, unlike the global intents that go to the previous part of the conversation.

Step 6 – Determine your variables and call to actions

Variables are placeholders that allow for a faster building process. A typical variable would be a phone number, email address, or contact us URL. You can also have full sentences which are often the same as a call-to-action such as “to learn more call 800-700-8077.”

Using a variable is simple, just type {{phone}} or {{email}} or {{contact us}} in the reply message box. When that intent is activated, the phone number or email address is automatically added to the reply message.

In the following examples, we show you how to create a chatbot using the Ometrics Ochatbot program.

Step 7 – Build your default question

Now you are ready to start building your chatbot. The following example is using the Ochatbot tool. First, determine how the chatbot will engage with the user. Do you want the default question to show first followed by a simple “how may I help you” on the home page? Another option is to ask a different default question on a different page or section of the site.

In the settings section, enter the introduction message. In the introduction message, you want to engage the user and inform them that they are speaking to a bot, not a human. This is important because you don’t want the user to be embarrassed that they thought they were talking to a person.

In the trigger section, select how you want the chatbot to trigger for desktop and mobile. Typically we suggest having the mobile not popup tand just stay in the bottom right corner as the chat bubble. This is because the chat window opening can be annoying due to it covering the entire mobile screen.

In the questions section, select the question you want to be the default question. The default question will show after the introduction text, depending on your settings. To switch to another question, assign the new question as default by clicking the checkbox.

Step 8 – Build other questions and conversational logic and intents

Below the question section, you can add specific intents that are associated with that specific question. Once intents are built, you can drag them over from the list on the left and edit them specifically for that question.

Step 9 – Build global intents

A global intent is an intent that listens all the time. It will trigger after the current conversation intents and before the assistant intent. When building these, think of the path that the chat user is going to be taking. Often these are connected to buttons and lead the user through a conversation.

Step 10 – Hitting the rails

Hitting the rails means that the bot does not pick up any listeners in the user’s response. A “fallback” is triggered when this happens to tell the chatbot user that it does not understand the question and it can provide the user a different way to get their questions answered. For example, the bot can respond with, “sorry I am just a chatbot. If you would like to contact a human send us a message or call us at 800-700-8077”. This can be an appropriate response at times because it helps the customer connect to the company.

Always test in real-time as you add your questions and intents.

For more help with chatbots download our Complete Guide to Chatbots and sign up for a free Ochatbot Account. Need help building a chatbot? Our chatbot consulting and development services can help. Click here for more info.

How to Build an AI Chatbot Free Course Signup

6 + 15 =