AI in the Customer Experience: The UPS, Downs, and Up-And-Coming Opportunities

However you might feel about it, a new age of artificial intelligence (AI) is here. Teachers have to double check that their AP world history essays were actually written by a student, bots can detect fraud better than a specialist, and consumers have gotten used to having arguments with their voice assistants. AI is causing disruptions for nearly every industry and business function— and customer experience (CX) is no exception.

AI is already playing a major role in customer experience in the form of chatbots, personalization, and interactive experiences. To find out how much of a role AI plays in the customer experience, we asked 2,201 consumers about their experiences with AI in CX, zeroing in on what they like, dislike, and would be interested in in the future.

Here are the top findings, plus advice for improving your customer experience with feedback— whether you decide to use AI or not.

Recommended AI News: MobileFuse Rolls out Outstream Capabilities and AI to Fusion Video, MobileFuse’s Video Suite

The chatbot takeover: Is it helping or hurting your business?

When it comes to customer support, there’s a pretty clear consensus; a resounding 90% of people prefer to work with a human over a chatbot. Across ages, genders, and income levels, almost everyone would rather “just talk to a person.”

Respondents said that humans: understand their needs better (61%), provide more thorough explanations (53%), are less likely to frustrate them (52%), and give them more options to address their problems (50%).

You might not be surprised by these findings—after all, who hasn’t found themselves in an infinite loop of dial pad options at least once—but the depth of the impact is striking. Not only is the consensus heavily in favor of humans, but that sentiment has a huge influence on how customers view the experience. Our research found that the average Net Promoter Score (NPS®) for human customer service engagement is +6 (slightly positive), while the average NPS for an AI-driven customer service engagement is -66 (very negative).

When we asked people to reflect on their most recent experience with customer service AI, here were some of the responses they shared:

It is extremely frustrating. Often, the AI system doesn’t include any option for my specific needs and it really makes me mad.

The chatbot kept answering a question that I didn’t ask, ignoring the question that I did—and then said it was ending our chat!!!

AI has no empathy or understanding.

When it is a straightforward issue, AI sometimes works. But when I have a complex issue, AI is impossible. There’s no context.

I have sometimes been taken around in automated circles by bots promising me answers, but actually simply referring me back to the same unproductive sequences. It’s like an infinite loop.

It’s also worth noting that usually when we include an open-ended question in a research survey, most people will skip it or answer in as few words as possible. For this question, we got many, many (relatively verbose) responses. If people feel passionately enough to vent about their experience in a survey, they might also be likely to do so in reviews, social media posts, or other public forums. That could be something to think about if you’re a growing brand that relies on positive word of mouth to build customer loyalty.

That isn’t to say that all AI—or even all customer service chatbots—are bad. Bots enable businesses to serve customers at more hours, more quickly, and often in whatever language they need. The takeaway isn’t that using bots for customer service is unacceptable; it’s just that human agents offer a meaningfully better experience today. That could also change as technology continues to improve.

Additionally, some people genuinely prefer to interact with AI. Among that group, the top reasons were better availability (41%), being able to address issues faster (37%), and access to more accurate information (30%). We didn’t ask about social anxiety, but it’s fair to say that that might have a role to play too.

Your success using chatbots and other AI might also depend on buyers and their needs. Only 41% of people under 34 have negative feelings about the idea of companies using AI in CX— compared to 72% of people over 65. If your customers are older, it’s especially important to have humans that are able to help them, even if that means slower responses.

AI and personalization: The next CX frontier

Chatbots might be the most visible aspect of AI in customer experience, but they’re far from the only use case. AI also comes up in product recommendations, personalized experiences, educational widgets, and more. Companies have started building and promoting AI-driven experiences, with apps that make recommendations based on an uploaded photo or the customer’s past preferences. And customers are more receptive to this type of AI when engaging with a brand.

Fifty-two percent of consumers are interested in having AI help them through a product, website, or feature experience. Forty-seven percent would like getting personalized deals based on their buying behavior—with another 42% interested in personalized product recommendations. That population could represent a fairly significant section of your market.

Callout stat: 52% of consumers are interested in AI that helps them through a product, website, feature experience, with 47% also interested in personalized deals and 42% in product recommendations.

Interest and excitement about these features are especially pronounced in younger generations. Among Gen Zers, 66% were interested in AI that would help them through a product or website, 63% liked the idea of personalized deals, and 56% wanted tailored product recommendations— a dramatic increase over the average response in every category. Businesses that want to connect to the younger, up-and-coming generation would be well served in incorporating other types of AI into their customer experience.

The use case matters here too. People, understandably, are much more comfortable outsourcing to artificial intelligence when the stakes are lower. Only 28% of respondents said they’d be comfortable using AI to access medical advice, and only 29% would turn to it when making investment decisions. But 65% would be comfortable using it to order food and drinks, and 59% would use it to return a purchase. Simple, transactional interactions with low stakes and less opportunity for ambiguity are where people feel the most comfortable putting their trust in AI.

Consumer’s perception of AI: hyper-visible but misunderstood

AI is certainly on consumers’ minds. According to our research, 89% of people believe that it will have an impact on their lives in the next five years, and 62% read about it in the news at least once a week. Yet despite the constant focus, AI remains confusing to many people, with only 18% saying they are “very confident” that they can tell when they’re interacting with a chatbot and only 14% feeling very confident about spotting AI-generated content. In fact, less than half (48%) felt confident they could identify AI-generated content at all.

So, it is possible that some people’s negative perceptions of chatbots might actually be about bad chatbots, and it is also possible that people are overlooking many other times that they’re already engaging with AI throughout the customer experience. It’s worth noting that there’s a generational divide here once again, with younger generations more confident in their ability to spot AI than older generations.

Recommended AI News: Mphasis, Kore.ai Partner to Transform Customer and Employee Experience for Enterprises

Use feedback to build a CX program that accurately reflects your customer preferences

You can’t ignore AI. It’s simply too powerful, diverse, and prevalent to dismiss outright forever. But you can definitely use customer insights to make informed decisions about whether to test new forms of AI, how to improve your current experience, and where you may have an opportunity to differentiate your brand from the competition.

Ask: Collect feedback while you test new AI

AI technology is constantly evolving, which means that companies that want to use it have to be regularly evaluating how it’s working. Maybe new developments will assuage people’s

complaints about chatbots and make them easier to use, or maybe consumers themselves will change, getting more used to working with their robotic support staff. Companies that want to stay competitive will want to experiment and see what works.

Collecting feedback as you go allows you to adapt when you need to. Sending a simple NPS® survey after each customer service engagement, for example, could tell you whether you needed to scale up your service team, make improvements to your chatbot, or focus your attention elsewhere.

Listen: Build feedback into the experience: One of the most consistent areas of frustration that respondents brought up about their AI experiences was being stuck in a loop where none of the options worked for them or the bot needed more context to answer their question. If you’re able to give customers an immediate outlet for their feedback, you’ll be able to help them faster and troubleshoot common problem areas simultaneously. It might also give them more peace of mind, knowing that if they can’t get their problem solved using the bot, at least they are still being listened to and valued.

You can build feedback options directly into your website or chatbot, or by triggering automatic emails after certain touchpoints (like a customer service interaction). Asking every customer to give feedback in a low-effort way, like a single-question survey, gives you a nice dataset to work with and improve upon.

Act: Use customer feedback to choose where to invest next. For some companies, an app with personalized recommendations for each customer could radically improve the customer experience. For others, it would be a fruitless drain on resources. Some interactions (like making returns) might be fine with a bare-bones chatbot, while others might call for something more sophisticated.

Proactively asking customers what they care about keeps you from making a costly mistake and can also help you understand which parts of your customer experience are making an impact.

Sending a customer satisfaction survey where you also ask about what would be most valuable to them gives you the answers you need to get your strategy right.

Like all technology, AI is a tool. Used correctly, it can bring you closer to your customers and deliver a better, more modern experience. You just need to be mindful about how you do it.

Net Promoter, Net Promoter Score, and NPS are trademarks of Satmetrix Systems, Inc., Bain & Company, Inc., and Fred Reichheld.

Recommended AI News: TELUS International Survey Reveals Customer Concerns About Bias in Generative AI

[To share your insights with us, please write to sghosh@martechseries.com]