Chatbot & Live Chat – A Perfect Match!

If you are you looking to move client or employee communication from the traditional (and more expensive) channels like phone and email to the cheaper and more efficient online chat options with Live Chat or a Chatbot, you may be sitting with the feeling that you need to choose between the two options. But do you really have to? I would say no – take both! Let’s have a look at why it can be a good idea to combine Live Chat and Chatbots into one single point of contact for your users.

Chatbot & Live Chat characteristics

Live Chat and Chatbots are two complementary ways of communicating with your end users. They shine on different characteristics and supplement the less strong side of the other in a very appealing manner.

Cost effectiveness

Both options are much more cost effective than phone or email channels where the support agent can only attend one person at a time. On average, Live Agents can engage in 3-4 conversations at a time. Chatbots, on their side, can engage in as many simultaneous conversations as the servers allow , and on the Teneo SaaS offering that capacity scales up automatically when the demand increases.

Availability

Humans Live Agents sleep, chatbots do not. You can of course scale up your Support Center to have Live Agents working around the clock, but that could potentially be an expensive affair. Unless you are in a field where 24/7 human support is a must, you can probably serve most inquiries in the non-central hours with the help of a chatbot.

Response time

Time is precious, and people want to get an answer to their questions fast. When chatting to a Live Agent you will get immediate responses, however, one of the main pain points for Live Chat is the waiting time before you get to talk to the Live Agent. Chatbots obviously shine on this aspect as they can handle many conversations simultaneously and will respond immediately to the request.

Human touch

The traditional phone channels are popular because humans like to talk to humans , and as an extension of that same behavior some people will prefer to chat to a human chat agent over talking to a chatbot. Chatbots on their side also have their advantages, such as providing fair and consistent responses. On top of that, conversations with chatbots can also be personalized and specific to what the user has said. With the right implementation, you can achieve that personalization that we all like.

Chatbot & Live Chat – a perfect match!

When you decide to implement both a Chatbot and Live Chat, you can take different implementation strategies:

  • Let the user chose if they want to talk to the Chatbot or the Live Chat Agent
  • Start the conversation in the Chatbot and inform the user about the possibility of being handed over to a Live Agent
  • Start the conversation in the Chatbot and offer handover to Live Chat when so defined in your conversation design .

Your strategy may evolve over time as your chatbot project matures, but no matter what strategy you start with there is a collection of scenarios where you should consider implementing the handover from the chatbot to a live chat agent when you go with a combined approach Chatbot + Live Chat. These scenarios can be divided into two groups:

  • Handover by Design
  • Handover for Error Handling

Let’s have a look at what we can find in those two buckets.

Handover by Design

In most Conversational AI projects, there will be situations where the bot cannot or should not handle the user query, such as

  • Low frequency - high complexity resolution processes where the effort of implementing the use case in the chatbot will not stand means with the usage it will get.
  • Questions that have not made their way into the chatbot, typically in the early stages of a project where the scope is more limited
  • Business critical situations such as complaints, cancellations of the services or outages with an impact on the business
  • Situations where sensitive personal data is required to respond to the query

Users may of course also explicitly ask to talk to a live agent, and the designed behavior can be to offer the handover as requested as mentioned above.

All these situations will not be present in all projects, and what is deemed critical or complex in one bot project may not be seen in the same way in another project. Some projects may also have implemented containment strategies to try to resolve the user requests in the chatbot channel which means that an explicit request to talk to a live agent may not lead to an immediate handover unless the user has previously tried to resolve the issue in the chatbot.

At the end of the day, the appropriate design of the user experience will always depend on the company who owns the bot and how they want to interact with their end users.

Handover for Error Handling

The group above lists the situations where the conversation design intentionally includes a chat handover because this is considered to be the happy path in those use cases. In other situations, the handover to a Live Chat Agent will be the endpoint of an unhappy path . Here we are talking about situations where:

  • A technical issue occurs, for instance with a backend integration
  • The user directly expresses frustration or anger (detected by the Teneo Sentiment and Intensity NL Analyzer, or when triggering one of the flows in the Conversational Module designed to pick up Negative Feedback)
  • Conversations run into a dead-end where the same flow is triggered repeatedly, indicating that the bot is not understanding the user correctly

If you have the option to handover to a Live Chat Agent in these cases, the likelihood that you can turn a potentially frustrated client into a neutral o maybe even happy one becomes a lot bigger.

Additional handover considerations

When you have determined your implementation strategy and the use cases in which you want to implement the handover to a Live Agent, there are a few additional considerations to take at different points of the process.

The answers to the above questions will be decisive for the end-to-end implementation and for the user experience.

Don’t chose – combine!

To summarize, if you want to implement Live Chat and Chatbot, there is no need to choose. You can go for both and let either type of chat communication do what it does best. Let the chatbot provide fast, consistent help 24/7 in the use cases that automate well, and back it up with Live Chat support in the cases where the chatbot cannot or should not help the user.

Chatbot & Live Chat are a perfect match and the combination of these two chat types will help you serve your customers in the best and most efficient way possible. The Teneo Engine API allows you to create a seamless handover by sending over the transcript of the entire dialogue or pieces of information from the conversation between the user and the chatbot, thereby providing the Live Chat Agent all the information that he/she may need to pick up and continue the conversation without having to ask the same questions.

Eager to add Live Chat handover to your chatbot? On Teneo.ai you will find a lot of resources that can help you get off to a quick start with the implementation of Live Chat handover in your Teneo Chatbot project, from sample solutions to sample connectors: Live Chat Integration.

4 Likes