As you join the race to integrate AI into your company’s buying journey, we all know the burning questions keeping you up at night: “How can I make my customers despise my AI?” “How can I create an experience so frustrating that they'll sprint to my competitors?” and “How can I make them feel so profoundly undervalued that they'll plaster the internet with scathing one-star reviews?”
It's no easy feat, mind you. Creating a truly loathsome AI experience requires dedication, persistence, and a complete disregard for everything your users actually want. You'll need to ignore decades of UX best practices, dismiss basic human psychology, and turn a blind eye to the very problems your AI was supposedly designed to solve.
But don't worry — we've got you covered! Follow these five tried-and-true strategies, and you'll have customers hate-Tweeting (hate-Xing?) about your AI in no time.
Ready to piss off your entire customer base? Let's dive in!
5 Ways to Make Your Customers Hate Your AI
We recently surveyed over a thousand consumers to learn what they like — and despise — most about AI. We compiled the findings in our new B2C Buyer Experience Report, so rest assured, the tips you’re about to read are backed by research!
Way 1: Force your customers to use it
According to our survey, 60% of people say they feel forced to use a brand's AI. And guess what? They’re not thrilled about it. Nothing says "we value your choices" like completely eliminating them! Perhaps the best way to guarantee customer frustration is to make your AI mandatory for every interaction. Hide the customer support phone number, remove the live chat option entirely, and make sure there's no possible way to bypass your digital assistant. You’ll have them screaming “speak to a human!” until their vocal cords are raw!
Way 2: Give it complex tasks that require nuance
Want to make steam come out of your customers’ ears? Easy: have your AI handle complex processes. A whopping 53% of our survey respondents said AI is worst at performing complicated, multi-step tasks that require nuance.
You can start by putting your AI on support calls where customers have interconnected issues. For example: "My debit card was compromised, so I need a replacement, but I also need to dispute three charges, and make sure my scheduled investment transfers don't bounce.” Watch as your AI confidently provides incorrect instructions, weaving a tangled web of follow-up problems that require three more calls to untangle.
Way 3: Give it tasks that require empathy
If those first two tips weren’t enough, consider this stat: 44% of people think AI is worst at providing empathy. So naturally, if you really want to irritate your customers, give your AI tasks that require emotional intelligence!
Picture this nightmare scenario: A homeowner calls your insurance company after their house was destroyed in a flood. They're devastated, exhausted, and desperately need help navigating their claim. However, when they explain the situation to your company’s AI, it responds with "I understand your frustration. Let me ask you a few quick questions to get started. First, have you tried removing the water from your home?" Tone-deaf experiences like these drive customers straight into the arms of competitors.
Way 4: Ask it to personalise conversations
While AI can give personalised Netflix and Spotify recommendations, there is a lot to be desired when it comes to contact centre interactions. A substantial 39% of survey respondents said AI is terrible at personalising conversations.
When your AI drops the ball on personalisation, it’s sure to make your customers’ blood boil. Nothing says "we value your relationship with our brand" like an AI that greets your customer of ten years with "Hi there! I see this is your first time visiting us. Would you like to hear about our new customer discount?"
Way 5: Count on your AI to perform follow-up tasks
30% of survey respondents were less than enthused with AI’s ability to follow through on requests. AI has the memory span of a goldfish, making it perfect for situations where customers need ongoing support! Nothing delights customers quite like explaining their entire situation from scratch every single time they interact with your "intelligent" system that somehow forgot everything from five minutes ago.
How to Make Customers Actually Like Your AI?
If for some strange reason you want customers to actually like your AI (weird thought, I know!), we have some tips for that, too. Our study found consumers like to use AI on their terms, rather than being forced to interact with it. They prefer using it for simple tasks where it can resolve issues faster than a human. Check out our complete findings in the graphic below:
Drive More Revenue with Invoca’s AI
Unlike the frustrating AI experiences we've been discussing, Invoca's AI is designed to create seamless online-to-offline buying journeys. Invoca can analyse contact centre conversations at scale, uncovering valuable trends and insights that help you improve agent training and customer service delivery.
This is AI that works behind the scenes to make human interactions better, not replace them with robotic alternatives — the kind of intelligent automation that customers actually want because it enhances rather than hinders their experience. Below is a sample agent scorecard, generated by Invoca’s AI:
In addition, Invoca data gives businesses a powerful foundation for training AI agents to deliver more human experiences. By using call transcripts, summaries, and conversation insights from high-scoring calls, brands can teach their AI tools what great interactions actually sound like. These real-life conversations capture the language, empathy, and nuance that drive successful outcomes. Feeding this data into your AI models helps them better understand intent, respond naturally, and mirror the behaviors of top-performing agents.
Download the Full Report for More Data
Want to learn more about how AI is transforming the customer experience? Download the full B2B Buyer Experience Report.
