Szczegóły lekcji

In this video we look closer at the chatbot automation modules. Learn how to set them up to create a properly-working chatbot.


Once we have our chatbot’s name and avatar set up, let’s have a look at the chatbot automation modules. I’m going to “Automations” and creating a new one. And then, among modules on the right-hand side we see there are 3 ones dedicated to chatbot: “Bot message”, “Answer condition” and “Assign agent”. In this video we’re going to focus on the first 2 modules as they are crucial ones in creating each chatbot automation path. “Bot message” is used to create a message text that will be sent to users by a chatbot. Similarly to module “Send a chat message”, we need to specify a trigger that will activate sending a bot message. It may for example be a pagevisit. But let’s look into details of the “Bot message” module. In the first field we specify message name, for example “Support question”, then message text so I’m typing in a short welcome text here. Then we have a few options related to chat messages in general - we need to decide whether to display a message as a new conversation or in the last active one and choose message type. Then there is one more field “Answer type” and thanks to the options available here you can adjust the answer input of the chat in a way that helps you collect correct (so in a correct format) data pieces about your users. In other words, thanks to defining an answer type here, your chatbot will know what type of input to expect. A very helpful answer type is “fixed” as thanks to it you can prompt 2 or 3 answers to a user to choose from. I’m adding 3 answers here “Software functionalities”, “Pricing”, “No help needed - I’m just browsing”. Then, as you see below, you can also save user answers to update their attributes directly from the conversation with a chatbot! It’s a great tool and we’ll cover all data types and a few examples in another video. Let’s now discuss the second module which is “Answer condition”. That module basically checks the answer a given user gave to a chatbot message and enables you to define content that should be displayed to users in the next step, depending on what they answered in the previous question. Important note: “Answer condition” module works always in connection to messages created in the module “Bot message”. They don’t necessarily need to be used next to each other and you can for example refer to bot messages from another automation in the module “Answer condition” but the module “Answer condition” has always a connection to a bot message. So for the “Answer condition” in the first field we need to select a message the module should refer to. In this case I’m selecting “Support question”. Then I need to select Operator and here we have “or” or “and” which are useful if you add more than 1 rule below. Let’s make a quick break from our “supporting bot” example now and see how the operators work. Let’s assume we have a bit more complex chatbot active on our website and its aim is to collect different data pieces about users. Among others, we want to know the age of people visiting our website as we have different offer for different age segments. We may go for 5 predefined age ranges as fixed answers and then set 2 answer conditions. If a user selects 1st or 2nd or 3rd - message A is sent. Then, if a user selects answer 4 or 5, they’ll see message B as an answer. So here we see how to set one replying rule for 3 different possibilites and it works if one of them is selected by a user. But let’s assume we don’t want predefined age ranges as we need to now the exact age of a user. We select “integer” as an answer type and then we can use “and” operator in specifying the answer condition. For example we want to send a specific reply to young adults so if an age provided by a user is greater than 18 AND at the same time lower than 30, this message will be sent. From a user perspective it looks like this. If I had typed in a higher number, another message (that in that case is dedicated to adults being 30 years old or older) is sent. So we know know how operator works. Let’s now go back to our main example of support chatbot. In general the “Answer condition” module is very often used with bot messages having fixed answer types, so here I select first answer and the second one and set the operator to “OR”. Then if the condition is met so, if a certain user answers “Software functionalities” OR “Pricing” I’ll can send the message number 1. In the second “Answer condition” module I’m setting a condition for third answer so “No support needed - I’m just browsing” and adding a different bot message as a reply. And if a user chooses this answer in a chat widget, a different message will be sent to them. As said, fixed answers are most commonly used but you can also set conditions for other answer types here. Last but not least, you may be wondering why this third module “Assign agent” is included in the green chatbot category on the right. Thanks to it you can assign one of your agents to a conversation started by a chatbot. Important note: a module is related to a conversation and NOT to a user so this is why it was put under the “Chatbot” category. It assigns a selected agent to the last active conversation of a user. So we see Elizabeth King here in the Conversations section, assigned to the conversation of a vistor who selected the “no help needed” answer. If you want to assign a user to a given agent, you can do it by a module “Update an attribute”, select the attribute “assigned to” and choose a preferred agent here. I hope we’ve cleared chatbot automation modules up, and as always - thanks for listening!