Customer relationship: which catbots to adopt and for which uses?

Three cases of use and implementation of chatbots within companies. Strategy, objectives and uses of chatbots in business.

The customer relationship is constantly changing and companies are constantly looking to renew their offers to consumers. One of the latest innovations in this area is to implement a chatbot to boost their customer relationship services and deliver an attractive promise of being available 7 days a week, 24 hours a day. In addition to being attractive, this technology also promises to save $ 11 billion annually by 2023 * for the distribution, banking and healthcare sectors. But for this mix of customers is really profitable, it is still necessary to understand the different cases of use of chatbots, and determine the appropriate one according to the needs and the specific objectives of a company.

The first mistake would be to consider that the use of a chatbot is unique and that it represents a universal solution to all problems. For example, two major Swedish banks, namely Nordnet and SEB, have associated the Amelia virtual assistant with their websites. Nordnet commissioned Amelia to accelerate the integration of new customers and improve their satisfaction. The results were not conclusive, Amelia was abandoned. As for SEB, her experience was more convincing, as Amelia had been used differently to achieve various goals. This proves that even the best chatbots can not always get good results unless its use is adapted to what a company wants to achieve. In fact, three cases of chatbot use can be set up within companies to meet specific needs.

A use for questions and answers

The simplest use case for chatbots is to answer basic and common questions from customers. These so-called "basic" chatbots, take the keywords contained in the questions of customers, and respond with answers already integrated and written in advance by human agents. In the end, these robots are programmed as interactive agents of FAQ sections. Their usefulness lies in their efficiency in intercepting rather simple requests to allow the physical agents of customer services to deal with more complex problems live. Simple to set up and use, these chatbots are also relatively affordable.

However, even though many companies start with this use, many of them want to quickly switch to more powerful forms of artificial intelligence. These scripted robots often represent a stalemate for the customers since over the discussions, the frustration settles because no truly conclusive answer is delivered. With the arrival of more and more sophisticated chatbots thanks to AI, customers can guess when they interact with a robot with scripted speech. Very quickly, they claim the help of a physical agent able to answer their requests, which proves the limits that these chatbots of use only for the questions-answers.

The use of conversational chatbots in a specific context

Ryan Lester, Director of Customer Engagement Technologies at LogMeIn

Ryan Lester, Director of Customer Engagement Technologies at LogMeIn

Companies investing in chatbots want agents who can go beyond the simple question-and-answer scheme. They want to bring value to their customers with robots capable of responding to more complex queries, and to have more human interactions where meaning and context are taken into account as responses progress. To do this, chatbots need to incorporate Natural Language Understanding (CLN) capabilities to leverage structured data and achieve smoother, more natural conversations with customers. Once these abilities are integrated, these chatbots are able to understand the intention of a sentence or an expression and to answer it more easily.

But to get a better idea of ​​the difference between an AI boosted chatbot and a pre-scripted chatbot, simple questions can be asked to distinguish them; and so the answers will vary as well. A scripted robot that is triggered with the question "Where is your nearest store?" "And therefore by the keyword" location "will likely respond by sending a link to a store locator tool, or even a list of stores where the customer will himself locate where the sign is located. A more advanced chatbot thanks to a superior artificial intelligence could respond with another question, thus pushing to really interact in this way: "Very well, by what means of transport do you want to go to this shop? And thus use the customer's response to guide him in his itinerary, and make this experience more useful and humane.

Agents capable of complex tasks and assisting with transactions

Using conversational AI robots like those mentioned above to accomplish tasks is the most complex scenario. For businesses, this requires significant financial and logistical investments, especially to link the chatbot to software and management systems. For example, if an airline-owned website wants to use a chatbot or a virtual assistant to book flights for its customers, the chatbot will have to connect with the entire network system in the background to assist in the verification. from customer identity, schedules, fares and miles systems to frequent flyers, all with great speed and safety. For companies that are considering more advanced use cases like this, it requires a lot of time and money. They must also ensure that the return on investment is worth it.

The potential of chatbots to improve, streamline and improve a company's customer experience is unlimited. However, it is necessary to identify what type of robot corresponds to the objectives of his company at the right moment, and which one may be able to adapt to the evolving expectations of a market in constant evolution.

Author: Ryan Lester, Director of Customer Engagement Technologies at LogMeIn


* Research conducted by Jupiner Research on chatbots, the results of which were published in July 2018.

Read the file: Artificial Intelligence and Marketing

(c) Ill. pixabay