Automated self-service: what separates failure from success

In times when speed of service has become critical and the concept of “normal business hours” has changed, we are becoming more comfortable to converse with technology like Chatbots to get our problems solved on our own terms. While some companies barely deliver a successful proof-of-concept, other companies build Chatbots that handle thousands of customer inquiries while delivering great customer experience.

Why do some companies fail to implement new technologies while others manage to transform their customer service and sustain high customer satisfaction?

Boost.ai, a leading conversational AI technology provider, invited a panel of experts to share their experiences on creating successful customer self-service at scale. I represented our fantastic team at Avo Consulting, one of the most experienced conversation AI implementation teams in the Nordics.

In this article, I summarize a few reflections about successful implementation of automated customer service channels like Chatbots. This advice works when implementing any kind of customer facing technology, so hopefully it will resonate with most of you.

You can also listen to the full recording of the webinar here.

Question:

There are many objections and concerns organizations may have when it comes to automation. Just because the technology has come so far that full automation of customer service is within reach, is it a wise course of action to let the machines do all the work?

How do you seek to find the right level between factors like cost reduction, automation and improved customer satisfaction?

Answer:

Some years ago, when we first experienced interest in massive automation using technologies such as Robotic Process Automation (RPA), many companies worried about how it would affect the jobs of their employees. We have yet to experience such negative effects of automation in customer service and most of the companies manage to reallocate their time efficiently. Today, however, the challenge is different, and it has to do with realizing value from automation initiatives. Many companies have become quite mature users of automation technologies, but they struggle to show value that their automation initiatives generate. This often results in discontinued projects and sometimes solutions that could be successful are cut off. Usually, there are two reasons for struggling to deliver on benefit realization (provided your technical solution is working as it should):

1.    Goals are unclear or not aligned

2.    KPIs for measuring goals are wrong

To define clear goals, you need to involve stakeholders who own the KPIs which the solution will deliver on. If you would like to create a new customer channel through a Chatbot, you need to make sure that business unit owners understand how the Chatbot can help them achieve their KPIs. Customer Service Managers will be focused on efficiency and customer satisfaction while Sales Executives will need to make sure that the Chatbot supports their sales KPIs. Involving the organization at different levels will ensure that you scope your solution in the direction that will be supported over time.

Another critical thing is to align the levels of KPIs with the nature of the solution in question. When it comes to Chatbots, it is prerequisite that they create great customer experience to be supported by the Customer Service Managers. However, measuring customer satisfaction based on the feedback you receive in the chat window can be misleading. Sometimes a customer will leave negative feedback simply because they did not get the answer they hoped for (like getting a refusal on a discount).

Also, decisions you make on the expected levels of customer satisfaction in each customer channel are important. Can you really expect a Chatbot to create the same level of customer satisfaction as your customer service agents generate on the phone? And if not, is Chatbot doomed to be a bad customer channel? Aligning expectations on expected customer satisfaction levels with stakeholders is critical to gain the long term commitment that any successful Chatbot initiave needs.

Question:

Automation and personalization have traditionally been considered mutually exclusive. But the increasing use of integrations and APIs that lets companies not only answer, but also solve inquiries relevant for each individual customer, changes the whole perception of these two being at odds. Do you believe automation will enable “mass personalization” for consumers in the near future?

Answer:

Personalization is simply about solving customer’s problem in the best way possible, whether it means with least steps, through personal service or self-service. We know from experience that customers use different channels for different types of inquiries. For example, if a customer wants to transact with us they will go to the website/app; if they want to validate their thoughts they will probably google, and if they want to discuss a topic, they will make a phone call. The way to achieve excellent service is to systematically use this kind of insight to match your customer’s preferred channels with the kind of service you provide in those channels. Designing good customer journeys is key.

Let me give you an example from one of the projects we have done. We know by experience that there are times when customers prefer to avoid contact with humans to solve their problem. An excellent example of this is a customer that is having trouble to pay his invoice within the due date. Calling customer service and admitting that you struggle to pay a bill can be a real strain. In this case a Chatbot is the perfect channel – customer logs in, is informed about invoices he can postpone and is helped to do so in an automated way. The customer can get help 24/7, without waiting in line or needing to admit the problem to anyone.

If you put yourself into your customers' shoes, you will be able to identify another opportunity in solving this kind of inquiry through Chatbot. The customer has such a strong incentive to extend the due date through the Chatbot that he is willing to pay a fee just to avoid the unpleasant call to the customer service. This means you can automate the inquiry away from the phone line, improve customer satisfaction AND generate additional revenue at the same time. This is the kind of value you can create through successfully combining channel strategy, customer insight and technology in the right ways.

The fact is that the state of Chatbot technology is mature enough to create personalized customer experience at great scale. The magic is in implementation and customer journey understanding, just like with everything else you develop towards your customers.