[Blog] 5 examples of how to ruin customer service with a chatbot
Providing great customer service has long been understood as central to the success of any business or organisation. From shop assistants providing helpful advice in person through to contact centres for phone and email contact, and, more recently, online through chatbots.
For these newer channels, quality customer service has never been more important to bridge the gap to customers through what can often end up being an impersonal level of service. Without the proper attention it is far too easy for these solutions to have the adverse effect and ruin a customer’s experience.
Principles of Good Customer Service
Customer service can be boiled down to several key principles that are considered essential to deliver a high-quality service:
- Access, speed & responsiveness – ensuring customers have access to a service that is responsive and meets their expectations in terms of speed. Speed of service can have a disproportionately detrimental effect on customer satisfaction.
- Accuracy & knowledge – providing answers and services that are high quality and contain accurate information to complete the service as required. Inaccurate information leads to the customer becoming frustrated.
- Clarity – delivering the customer service in a clear and concise way. A well intentioned process that confuses a customer can only lead to reduced satisfaction levels.
- Transparency & honesty – keeping the customer informed on what is happening regarding their request and the service. Being slow to provide updates is frustrating for customers and leads them to assume that there is some issue you don’t want to tell them about. Honesty is essential even if this means admitting that there is a problem.
- Empowerment – giving the customer the knowledge or feeling that they are in control. This helps build confidence in the service because the customer knows that action can be taken if they are not happy.
Obviously, there is some overlap between some of these areas. For example, speed of service is often dependent on those delivering the service having the required knowledge/information about the customer.
The importance of these principles for good quality customer service is clear. This is also true when providing customer service online where it is, in many ways, more important. Online customers have more choice, higher expectations and the ability to switch to a different option with ease.
Robot Customer Service
Online, or digital, channels also offer many benefits. They allow organisations to reach far more customers and offer services and quality that are not necessarily constrained by their size or capacity.
Digital channels also facilitate the use of innovative technologies such as Artificial Intelligence (AI) and Robotic Process Automation (RPA) to perform specific functions. For example, AI driven marketing and chatbots, that can be used replicate functions traditionally performed by humans to deliver them at a far greater scale.
Chatbots are becoming integral to providing customer service online. These are not to be confused with live chat where there is a human contact centre agent engaging with the customer through the chat software.
Chatbots engage the customer without the involvement of a human contact centre agent. They use AI, rules, machine learning and automation to understand the customer intent and provide responses to support them meet their needs. The services that a chatbot can provide are usually pre-designed and therefore it should be possible to create these with customer service principles in mind.
When robots go bad
Despite this there are many examples where chatbots fail to deliver and customers are left frustrated. Here are some examples:
- No response – The customer engages the chatbot, possibly asks a question, and then nothing happens. Or worse, a spinning circle indicates that something is happening, but no response appears. Clearly this impacts the customer service principles of responsiveness and speed but also leaves the customer feeling helpless and frustrated.
- Pointless – The chatbot responds, asks meaningful questions, and appears to be helping only for the end response to be that it cannot meet the request and the customer should call the contact centre. Quite often customers will go back and try a different combination of answers to try and find an alternate outcome. This results in customer frustration due to the lack of clarity, knowledge and possibly even trusting the service’s
- Chatty – Advanced chatbots use AI tools to communicate with customers in their own language rather than asking questions and providing options. Although this can be a powerful tool if customers are not clear on what service they require it can result in service issues. If the chatbot cannot determine the customer’s intent it will keep asking questions or, worse, will tell the customer it does not understand. Significantly impacting the customer experience due to the lack of clarity and accuracy and creating an overall delay to service delivery.
- Lists – Some chatbots are based on functionality that searches the organisation’s Customer Service databases. The customer types in a question and the words are used to find the topic, which is presented back to the customer. In the worst cases this is a list of pages or options for the customer and they are expected to search through and find the answer. The lack of accuracy, knowledge and clarity all severely impact the customer’s experience.
What We Think
If these issues or the bad chatbots sound familiar, then you can do something about it! Do not underestimate the potential impact of a poorly performing chatbot, even if your customers have other options for interacting with you. They may leave and not return. It can also drive an increase in contact centre call volumes as customers still need help or want to complain.
Furthermore, if the customer in questions is online using a chatbot then they are likely to be happy to vocalise their frustration on a company’s social media channels. Resulting in their experience being shared with many more customers.
Make sure your chatbot is built to do what is supposed to do. Do not be overly ambitious with the breadth and scale of options to offered initially – the priority is to provide a service to customers and build on that. This might be as simple as some questions and several potential answers but if it meets the customer requirements in a faster, more responsive, and transparent way then it’s a great first step.
Take the empowerment principle seriously. If possible, offer the customer an option for the chatbot conversation to be passed to a human counterpart if it cannot find the solution. Even better, use AI tools to monitor sentiment and act if the customer becomes frustrated.
Despite these examples, chatbots, when designed and built well, can offer significant benefits. They can engage customers to provide the simple services, reducing the demand on contact centres, releasing capacity to focus on more complex issues.
A customer who has a good experience with a chatbot is likely to return and use that service again. In parallel, chatbot capabilities can be built up over time so that they take on more and more services as and when they and their customers are ready.