Managing conversations
Last updated
Was this helpful?
Last updated
Was this helpful?
If you are a customer care leader, business owner, or project manager, one of your main goals is to make the customer communication process as smooth as possible. Activechat can help you achieve this in a number of different ways, which we'll explore in this section in detail.
First of all, let's briefly review the iterative nature of customer service automation. This process consists of multiple stages, which you can find in the illustration below.
Customer service automation workflow is a constant process of improvement that helps you build better experiences for your clients and free up valuable human resources for your team.
Key components in this workflow are listed below (in no specific order, since you can start the customer service automation process anywhere depending on your current situation):
For most companies, every new customer interaction starts with an inbound request (a message in the chat, an e-mail, or a phone call), which is then handled by some SOP (Standard Operating Procedure), and the response is sent manually by an agent. This process can take some time, and customers hate waiting.
You can improve this instantly with our live chat automation platform, and here's a brief overview of what the process looks like (both for your customer and for the support agent).
While customer care automation can definitely improve your team performance, in no way should it compromise the level of personalization that your customers get.
Activechat takes this personalization to the next level, allowing you not only to use traditional customer attributes like name, email, or phone number but track any kind of information that you feel important down to every single interaction with your business (like visiting a specific website page or adding item to a shopping cart).
This information is not only stored in your live chat CRM but can also be used to personalize communication with a customer.
Talk to your customers. Get actual conversations done to understand your customers' pain points and where they need support most. Our can help you do these conversations in the most efficient way.
Get conversational AI insights. Understand which phrases your customers use to describe their issues and how these issues can be grouped into specific topics. This information is available in the , and you can use it to create new automation intents and update existing ones for better virtual assistant performance.
Create new intents and update existing ones. It's extremely important to keep the conversational AI and knowledge base data well-organized, and that's where our can help.
Automate repetitive tasks. Build interactive conversational scenarios that will guide your customers through the process of solving their issues or getting the information they need. These conversational automation scenarios can easily be built with our .
A new message is received from the customer through one of the connected to your project in Activechat.
This message is analyzed by the virtual assistant first (though the ) and an attempt to match the message to one of your existing (skills) is made. We utilize state-of-the-art natural language understanding (NLU) technology to extract the meaning of the message even if it's misspelled or put in different words.
If the match is found, appropriate action is taken. It can be as simple as sending back the pre-defined response or as complex as running a user through a series of questions to help solve the issue or get the necessary data (for example, for purposes).
If the match is not found, the bot can respond with something like "Sorry, I do not understand" (which is obviously a bad user experience), suggest some generic help, or (which is much better) connect your customer to a human agent. You can use to assign the conversation to the most appropriate team or person.
Once the live chat session is done, the agent can , and connect it to an so that similar requests will be processed automatically in the future.
An alternative approach could be using live chat to handle all incoming conversations first (maybe with some help of our ), and use our feature to extract valuable data from your customer conversations. To set up this kind of behavior, check the guide in the .
Every skill that your virtual agent has can be customized by using the block in the visual flow builder. This block will branch the conversation depending on the value of some specific customer attribute (like the number of orders or account type, for example).
