Trending Articles
Trending Topics
Trending
This article is part of our Opinions section, where we invite industry professionals to share their views on the most pressing technology questions of our time.
AI is in everything now, and we all find ourselves interacting with it on a daily basis. When it comes to productivity, making ourselves do more in less time, that’s great, but when companies let their customer interactions be taken over by chatbots, they may be in some trouble.
You will have noticed by now that when you reach out to companies online, the responses seem canned somehow. The customer service rep never seems to quite answer your question, or instructions on how to navigate the company knowledge base come a little too quickly.
This is either because you’ve dealt with a rep who’s very good at navigating through a database but not so great at talking to people, or, more likely, because you’ve been talking to an AI. Chatbots are a way of life, and there’s no way to get around them, at least on first contact. The question for businesses remains how wise it is to implement them this much.
Using AI for customer contact makes sense from a financial perspective. Humans are expensive, and many customer concerns can be handled quickly and easily. If you’ve ever had a customer-facing role, you know this: most of the stuff people approach you with can be handled quickly and with just a few sentences.
It seems like a waste to spend money just to answer really basic stuff: “How much is the beginner plan? How long does a subscription last?”. And it is, and many businesses have struggled with ways this cost could be cut, or at least minimised.
This is where AI comes in, which, on paper, should be able to handle the basic stuff. A chatbot can handle first contact with any customer and either answer the question or send anything too complicated to a human. Recent AI systems that have hyper-personalisation built in can even tailor responses and recommendations based on customer history, preferences, and sentiment analysis.
When all this is applied well, this can make even first-line AI responses feel more human and relevant, increasing customer satisfaction. This should mean that on the one hand, you need fewer people manning the proverbial phones, and that their time is spent on more interesting questions. A win-win for everybody.
Well, a win-win except for the customers, who now need to deal with AI. If you’ve ever been in that situation, you know how bad it can be. The AI’s limited input options mean it will likely misunderstand what you’re asking unless you nail the question perfectly, and any time you don’t do so, you’re punished with the same handful of links to a labyrinthine knowledge base.
While plenty of pioneers are working on improving the emotional intelligence of AI and increasing its capacity to detect frustration, confusion, or urgency in a customer’s tone or language, the tech just isn’t there yet. Right now, even technically correct answers can feel dismissive or robotic, eroding trust in the brand.
It’s bad, and it’s getting worse as the headline – AI will save you money on support staff – drowns out all the caveats that come with it. Add to that how the designers of these systems always seem to assume customers know perfectly what they need from you, and there’s barely any pushback against the widespread use of AI.
But all the assurances of AI’s salesmen won’t change the truth: customers hate chatbots. Gartner published a survey late last year that said 64% of respondents would prefer companies not use chatbots, and I have a sneaking suspicion that number’s only grown if I go by all the complaining I hear from clients and online. In fact, the only people who seem to think AI chatbots are a win are the people selling them.
The thing is, the attributes that make AI great for increasing productivity – figuring out a simple situation and then acting to fix it according to pre-set templates – are exactly what you don’t need when dealing with customers. Humans are all unique, and so are their issues and use cases. If you treat them as one homogeneous mass, you’re doing them a disservice. You’ll also make them angry.
We knew this even before AI came around, when call centre operators were forced to follow a script; AI is the same situation, writ large. The weird thing is that the scripts and AIs themselves are made to sound as warm and human as possible – everybody knows you need the human touch, just nobody wants to spend money on actual humans.
While AI can be a powerful tool, relying on it too much risks distancing you from the people you serve. A balanced approach could offer a solution for now, one which incorporates innovations like multimodal AI to respond more accurately and quickly to human queries and can pass issues on to humans more easily.
In the end, there’s no compensating for the human touch, and sooner rather than later, this will show in which companies are successful and which aren’t. In any market where customers have a choice, consumers will move where they are treated as humans, not greeted by a robot.
The companies that excel will blend the efficiency of AI with the empathy of humans. For example, letting AI handle routine queries, but empowering human agents to step in promptly when complexity, nuance, or emotion are at play. In this future, AI isn’t the replacement for human service; it’s the support system that allows humans to shine.
I predict that as AI takes greater root in our society, this backlash against an AI-only approach will only grow. If you want to stand out for your customer service as a company, you’re going to have to put people first, and you’ll need people to do that. AI, for all its strengths, just can’t do that, and probably never will.
