11 Jul 2019 - 5 min

Chatbots: the quest for ROI (Part 2)

Why properly designed chatbots are still a tool of the future to provide the best possible customer experience?

In the first part of this blog post we looked back on the unfulfilled promises of chatbots and touched upon the fantasies that still surround machine learning (ML) in 2019. In this article, we will, among other things, explain why we are convinced that properly designed chatbots are still a tool of the future.

 

According to Gartner, 25% of customer services will use chatbots next year. Markess tells us that 66% of professionals will have tested or deployed conversational agents within two years. For the 88% of companies that make customer experience a priority, it is therefore crucial that they equip themselves with a tool that complements digital experiences!

But then, what should the chatbot’s look and feel be like? According to the Capgemini study, “The Secret to Winning Customers’ Hearts with AI,” 73% of consumers say that they know when they interact with artificial intelligence. 64% would like to interact with a “more human” artificial intelligence whose voice is less metallic, and with the ability to conduct a conversation, that is, to answer a series of questions. At the same time, with the progress leading to anthropomorphism, 2/3 of consumers want to know who they are dealing with – man or machine, virtual agent or real agent.

Firstly, the chatbot must find its audience! And before finding the right location (website, application, social networks, voice speakers, phone, etc.), the chatbot must first find its mission. It is better to start “small” by addressing a single issue (meeting the need for client autonomy, relieving agents, etc.), before thinking of addressing several.

 

 

What technical, human, and financial resources does the company have at its disposal to build, promote, and support its chatbot? These are all questions that need to be answered in order to provide the best possible customer experience.

 

And the measurement of ROI in all this?

 

There are many performance indicators (KPIs) to measure the effectiveness of your chatbot. If the Net Promoter Score (NPS) provides an overview of the customer satisfaction measure, reuse rates will immediately indicate whether the chatbot is a success. In addition, a high first contact resolution rate will contribute to long-term adoption. Finally, measuring the effectiveness of the system in relation to its initial objective is crucial, and being able to compare the use rate of this new channel with other channels is just as important.

This leads directly to economic performance assessments: the number of cases handled outside of and on the hotline, increase in self-service, reduction in waiting time online, etc. Over time, it will also be interesting to monitor the number of users, the number of use cases addressed by the chatbot, the time spent with the chatbot, the contact with an agent following the use of the chatbot, or the reduction in processing time per agent following the qualification work performed by the chatbot.

Finally, chatbots are also intended to secure sales: the number of secured baskets or the number of hot leads detected and immediately transferred to a human agent to prevent them from disconnecting, are KPIs that need to be considered. However, keep in mind that the “good” KPI varies from one company to another, from one use case to another, even from one chatbot to another.

 

To the ideal chatbot and beyond!

 

In recent years, chatbots never ceased to be a conversation topic. Seduced by the promise of increasingly humanised robots, customers are excited but also anxious. In order to win them over, companies must succeed in setting up scalable devices that meet their customers’ needs, are able to help their agents, and whose ROI can easily be demonstrated.

Does it seem like mission impossible? Not with the Odigo solution. We understand that to build a chatbot that provides the best services for its customers, companies need to make sure that it always addresses relevant use cases. We also take the success factors of a chatbot into account to meet the customer’s ROI objectives:

  • Its ability to understand natural language needs to fuel an engaging interaction.
  • Its ability to provide qualification needs to prepare the agent and reduce his workload as much as possible.
  • The automation it offers needs to meet customer self-service challenges.

 

In agreement with your teams, we also give the chatbot its own personality. Finally, we make sure to offer the best user experience (UX) by paying close attention to ergonomics.

The use of natural language makes interactions more user-friendly and effective. Designed with the entire journey in mind, our chatbot intervenes on the channels used by your customers and delivers a personalised service using a question-and-answer database and/or accessible information. Because it is fully integrated with the Odigo solution, it is easy to transfer the discussion from the chatbot to a real agent when requested by the customer.

 

Interested? Request your demo here.

 

 

chatbot
customer experience
machine learning
self-service