Are chatbots truly the future of customer service facilities?

Recent months have really hyped up the role of “bots” in brand communication, specifically in the customer service function. Nobody is surprised. Customer service has always been an area where brands, quite illogically, focus almost entirely on cost-cutting. While headlines of press releases from the likes of telecom and real-estate companies would have you believe that their investments are single-mindedly focused on customer experience, the reality is very different. From incentivising shorter contact centre calls to the new hype around chatbots, customer service is a lot more about cost-cutting than brands want you to believe.

Image: Shutterstock

Chatbot adoption is still a whole lot of optics

The conversations and PR spins applied to chatbot adoption have a lot to do with “optics” (perceptions) of the brand, such as wanting to appear “cutting-edge” and showcasing first-mover advantage. But after several “conversations” with these chatbots, I have come to realise that they inherently mirror the inefficiencies of customer service in general, with their robotic drawer statements and zero human touch. I am yet to come across a single brand’s chat bot that gets it right as far as problem solving and natural conversation are concerned.

To understand the hype and separate the wheat from the chaff, we must first understand that a chatbot really is just another platform. The premise is still one of servicing customers’ individual needs and questions and providing quick solutions.

But current chatbots are not human – at all

Most evolved customers will interact with a brand only if the brand fails to deliver self-service in a way that works. When a customer eventually picks up the phone or opens a messaging app to speak to the brand, he/she is already frustrated. If chatbots respond, as they usually do, with contact centre numbers or drawer statements, what purpose do they have? Aren’t they just a waste of an organisation’s resources because they just state the obvious, and in the process, increase the number of steps the customer must take to eventually find the answer to their question?

A lot of work has gone into making chatbots intelligent. But the bottom-line is that bots operate on logic trees and respond to keywords. They almost never respond to natural conversation. Take Siri for example – she is still struggling with my very Indian accent!

Complex problems are still a no-go

I will give you a recent example. I was trying to find a general physician at a popular hospital in Bengaluru. The website did not list any general physicians, and that’s when the chatbot appeared from nowhere and asked if it could help. I asked my question, and in response, got the website link where I could find all the doctors, and a phone number I could call on to book my appointment. If my need was as simple as a website link, I wouldn’t need a chat bot. Most customers wouldn’t. The only time customers start a one-on-one conversation with a brand, either on social media or on a messaging app, is when their problem is more complex. If chat bots can’t respond to such questions, I don’t understand the need/purpose they actually fulfil.

But can chatbots actually work?

Sure, they can. But for that, it is important for brands to understand that their bots, along with their other customer service platforms, need to evolve into quick problem-solvers. Whether they deliver it through a messaging app, their Twitter handle, or their contact centre, is just a matter of platforms. It is not something customers would care deeply about as long as their questions are answered and their problems solved.

The lack of human touch in bots can be a strength or a weakness. When customers call a contact centre, they often have to repeat themselves, paraphrase their problems, or speak in an accent that is better understood in the region or country. Chatbots can eliminate these dependencies. Good, functional chatbots will be the ones that are trained to resolve customer issues at first contact without letting grammar and accent come in the way.

Artificial intelligence must get real

Bots must be conditioned to understand individual nuances of customers over time. This will ensure that that they respond to each customer in a way that is personalised and feels natural. This is not always possible with human customer service associates catering to extremely high customer volumes. Until bots become the answer to limited human intelligence and memory, they will not be able to find a purpose in the customer service ecosystem, which runs more on human emotions than most other business functions.

Bots need to be designed to be faster and easier to use than making that call to the contact centre. That is the only way for them to become more utilitarian. It is important to apply the unique concepts of human computer interface systems into chatbots. They need to be more intuitive and respond to fewer clicks, touches, and typing. Chatbots must have pre-built options based on realistic customer queries and issues in order to be useful. In essence, this is similar to contact centre IVR or FAQ manuals.

Chatbots do have the capability to revolutionise customer service as we know it. But in its current form, the technology is far from delivering anything consequential. Before we write off or reward chat bots, let’s wait and watch how they evolve over time, and how quickly brands realise that chat bots or any new technology will not work till the time their customer service strategy itself is not built on the tenets of human empathy. New technologies and platforms don’t automatically solve old, complex problems. It takes intelligent people to train and tame these technologies and make them more human in order to be effective and deliver actual results.

Read Also: Everything you need to know about chatbots