October 2, 2024
©PhonlamaiPhoto via Canva.com
Should Consumers Be Anxious About AI’s Impact on Customer Service?
A new survey finds that consumers see AI worsening customer service, primarily due to concerns over data use as well as an absence of empathy tied to the lack of human interaction.
The survey of 1,520 U.S. adults from real-estate services firm Colliers found that 54.3% of consumers expect overall customer service at retail to worsen due to AI’s impact, with only 13.2% expecting the technology to improve customer service. The remaining 28.2% expect customer service to remain the same.
When asked what specific concerns they have about using AI in customer service, the top responses were:
- How my data/information is used, cited by 62.7%
- Lack of empathy from AI, 61.3%
- Lack of human interaction, 54.5%
- AI not properly understanding, 54.1%
- Inability of AI to resolve problems, 47.6%
- Lack of friendliness, 46.6%
- Questions/queries taking longer to resolve, 41.3%
Also, 67.2% felt AI will have a “worse” impact on retail jobs as roles are automated. On the positive side, most consumers saw a positive impact from AI on product availability (41.4% “better”/18.9% “worse”), website experience (46% “better”/ 12.4% “worse”), and promotions and offers (36.4% “better”/18.9% “worse”).
An accompanying survey of 59 U.S. retailers found just over three-quarters are optimistic about using AI. The top initial usage areas were product recommendations, day-to-day data analysis, generating website listings, supply chain management and logistics, and forecasting demand.
There is still hesitation when it comes to customer service and support, with fewer than 50% of retailers anticipating the adoption of AI in these areas. Similar to the consumer survey, the top negative impact of the use of AI was increased data security concerns, cited by 39% of retail respondents.
Colliers said the study shows retailers need to do more to highlight the benefits of AI, including being able to free up retail associates from tasks like inventory management and floor merchandising to deliver both elevated and more personalized customer service. Colliers wrote in the study, “This shift promises a more fulfilling and engaging work environment that maintains a genuine connection between the brand and the consumer.”
Capgemini’s 2024 edition of its “What Matters to Today’s Consumer” study found that consumers who have used AI for shopping have a positive outlook on GenAI tools, particularly for product recommendations. However, extensive media coverage has led to a sharp uptick in concerns about GenAI producing false/misleading testimonials or reviews as well as the technology’s capability to copy or clone competitors’ product designs or formulas.
The study found that 47% of Gen Z and Millennials would like a chatbot with features such as ChatGPT to facilitate asking questions and receiving responses quickly. Nonetheless, 70% of overall respondents who have used GenAI for shopping appreciate prior disclosure if their interaction will involve GenAI, and the same 70% want an option for a human representative in case GenAI’s assistance is unsatisfactory.
BrainTrust
Shep Hyken
Chief Amazement Officer, Shepard Presentations, LLC
Shannon Flanagan
VP|GM Retail & Consumer Goods at Talkdesk
John Karolefski
Editor-in-Chief, CPGmatters
Recent Discussions
Discussion Questions
Are consumers’ concerns about the potential negatives and risks around AI-supported customer service legit?
What should retailers do to ease concerns or overcome any shortcomings?
Poll
Loading …
This research was done by my firm, GlobalData, for Colliers. It shows a real mismatch between retailers and consumers when it comes to the impact of AI. There is a general perception among consumers that retailers will use AI to cut corners and this will lead to worse experiences, especially in terms of customer service. Whether this comes to pass or not remains to be seen. However, where we have surveyed consumers about their experience with AI chatbots the majority are negative. It’s not that consumers object to AI per se. It’s that they want to have a choice and that choice has to involve being able to speak to a human being if needed. The future is blended: AI will play a role but it won’t replace the human touch.
If a retailer isn’t going to train and retain human beings, who are infinitely adaptable, can ask clarifying questions, and give practical advice based on firsthand use of the company’s products, then the retailer sure as heck isn’t going to invest even more into testing and refining technology that has nothing to do with its core business that would allow it to offer a superior customer experience. Rather, the retailer will go for the cheap option of a gussied-up predictive text phone tree, lay off staff, and tout this on its investor deck while giving bonuses to executives. We all know how this cycle goes; the consultants make good money and the whole exercise gets canned after 18 months.
Can it get worse? I think the mistrust comes from companies like Xfinity claiming they use ai to deliver the absolute worst customer service experience ever. Or some of the big banks. Or Facebook, that randomly puts people like me in “jail” because it’s programs are purely keyword driven with no sense of context or irony.
All one can think of is “ if this is ai I want nothing to do with it. Now, the truth is, there is no ai used anywhere in retail except in the supply chain, and that’s more machine learning than pure ai.
The hype is kicking those who proselytize the technology as the savior in the tush.
This. AI is not for customer service. I spend more time explaining what I want than explaining my needs to an actual person. i just don’t think people want cold and impersonal and won’t for a long time to come.
When digital customer service, fueled by AI, is done right, the experience can be better. The problem is some get it right and some don’t, so customers aren’t getting the consistent experience they want, which impacts their overall view of AI-supported customer service.
My annual customer service and CX research found that only 32% have successfully resolved a customer service issue using AI without human assistance.
There is no doubt that artificial intelligence is a monster from both heaven and hell. There is no way to tell at this point in time whether it will be more of one or the other. At present, AI is at the same stage as the internet was in the mid 1980s. While most of us recognized its high technology, its potential greatness, and its potential to change life as we knew it, our imaginations took us to both good and bad places.
For retailers or consumers, it is way too early at this stage to fully understand, appreciate, or despise features, benefits, and drawbacks, of which there might be too many, at least until the dust settles, if that ever occurs.
The initial effect of AI on customer service will be mostly positive, but as time progresses, there will be “issues” that will be hard to resolve. I think we can all agree the same has been true with the internet. Db
Let’s get one fact out of the way: many of the latest AI capabilities are not being built in-house by retailers, they are provided by third parties. That’s good news and it changes the conversation. The innovative and capable customer-service-enabling third-party solution providers that I’ve spoken to just over the last month have restored my faith in the AI-driven future. These companies have a vested interest in providing security, constant innovation/iteration, training, and service after the sale if only because they already face so much competition. In this case, reality is more powerful than perception.
Another reality is that these third party offerings are only as capable as the data which is provided to them by the retailer. Without well organized and managed data, the customer experience is suboptimal regardless of the software.
Absolutely true and thankfully, I hear that is getting better by the day for software that relies on retailer-supplied data. Some of the most exciting solutions grab data directly from shoppers. The ability to track, quantify, and act upon in-store activity is particularly exciting these days.
My hunch is that many consumers mistrust the potential for AI service bots based on their abysmal experiences interacting with dumb-as-a-rock chatbots and infuriating IVR phone systems. Both are programmed with human-designed algorithms that anticipate a very narrow set of generic service issues. As a result they usually thwart the delivery of quality service, not enable it.
As I understand things, a true artificial learning intelligence could likely be “trainable” to deliver better service experiences, but the training period will be probably be excruciating for humans who interact with it.
Until they can understand and reply helpfully to plain-language queries from naive customers, AI service bots will deliver more frustration than satisfaction.
Retailers who aspire to adopt AI customer service should think twice and test carefully. This may be one of those innovations where it is far better to be a “fast follower” than a “pioneer.”
When the AI-driven experience is all about forcing the customer to figure out how to get around the AI so they can get to someone who can actually help them, then consumers are definitely going to hate it. Not to mention the general cynicism that no savings will be invested in anything for the customer’s benefit or for the benefit of the employees who remain.
I know I find that I am reaching out to a brand because I have a problem that is more difficult to solve than can be answered by an FAQ – and by extension, a chatbot. Respect customers and give them an exit hatch. Or at least, analyze your logs to stop endless loops from happening.
One last thing: my uncle passed away in July, and I am helping my aunt close all of his accounts. You want to judge a company’s empathy for customers and the sophistication of their AI? Go through trying to tell a brand their customer is dead and see what happens next.
Amen sister, though I’m sorry about your uncle!
Well implemented AI should be a significant benefit for companies and customers. The “Well implemented” part is critical.
It’s still early days with few AI practitioners and little AI practice. Using 3rd parties with more and specific AI experience helps things get started, but AI skills need to be developed in house to ensure a high quality customer experience.
The wins for companies and customers look a lot alike. Companies can train the AI once and offer customers consistency across responses. No differences rep by rep to confound customers and result in more issues for companies. More AI agents can be made available to limit wait times for customers and improve the customer experience at nominal cost.
Expert human agents should always be an option for customers and the AI to select. But AI should be capable of solving a high percentage of the most common customer service issues quickly and accurately.
Periodic updates based on interactions keeps the AI “trained” up to the minute.
All of us have had infuriating experiences that leave us shouting “OPERATOR!!” into the phone.Let’s hope that the companies that implement thee technologies test them adequately, and provide customers with an alternative if necessary.
Yes. Consumers should be concerned about AI’s emerging presence in customer service as it is yet another downhill step in respect for the humanity of doing business. We are, after all, in a business of humans. The primary reasons given for AI are generally those of thriftiness — the eternal search for efficiency. Retailers must come to realize that an efficiency driven focus a race to the bottom — poor customer service, bad products, and more. The ONLY path to business health is one of effectiveness — bringing more people to the store, keeping them coming, and satisfying them (in general). That is a matter of product and experience. There is no indication AI helps a company take a step forward toward healthy companies.
For many years retailers have been using customer date to personalise offers. There has been a deal struck whereby customers share their personal data with retail brands they trust and in return retailers provide relevant and timely offers or positioning of products to customers.
AI is likely to add to the precision of these offers – from some respects this increases the convenience for the consumer because the offers or positionings become even more relevant.
AI also has a role to play with customer service – providing the ability for consumers to have their queries and issues fixed faster. Some may worry about ‘talking to a machine’ but if queries are fixed seamlessly this concern will be outweighed.
To give consumers peace of mind, retailers can always offer the ability to ‘talk to a human’ and indeed many of the uses of AI are focused on freeing employee time for them to spend more time with customers.
Consumers’ concerns about the use of AI for customer service are legitimate. Chatbots that have been leveraging AI for several years have been a disappointing and frustrating experience for savvy consumers. Personally, before I engage with a chatbot, I research the retailer’s website to answer my questions. I explore every option and by the time I resort to help, I probably know more than the chatbots. I have never had a problem solved by a chatbot and I just continue to request a “live agent” to get a human to solve the problem. I am still not convinced that AI is best suited for customer service, unless it is used to direct you to the right human.
The most promising areas for AI and generative AI are in improving operational efficiency, product recommendations and marketing.
It is too early to tell for sure what will happen. But consumer concerns should always worry retailers.
Customer facing AI in service to date has sucked, no doubt about it. But it’s because the technology has been immature and companies are over-indexing on “call deflection” as a key KPI. As LLMs become more powerful and human-like, service may be better than ever with genAI, just give it time.
AI for customer service most likely will follow a rules based system and it will be every bit as frustrating as listening to “for store address press one” and so on-worthless.
In some ways putting “real people” back on the phone is worse, because the business model dictates that cost trumps everything, and that means low wage workers following a script. Tedious and unhelpful.
AI is going to happen whether we want it to or not. The real issue is how much horrible customer service will consumers tolerate, and I believe the answer is a lot-primarily because this aspect of retail is a race to the bottom.
More Discussions
How Important Is ‘Vibe’ in Physical Retail?
Not attempting to be too playfully groovy about this question: According to a recent report from…Will Pinterest Translate to Shoppable TV?
Will Pinterest lead the way in shoppable TV?How Can Retailers Improve Shoppers’ Sense of Closure When Concerns Emerge?
How can retailers better provide closure and proper resolution to today’s shoppers?
This research was done by my firm, GlobalData, for Colliers. It shows a real mismatch between retailers and consumers when it comes to the impact of AI. There is a general perception among consumers that retailers will use AI to cut corners and this will lead to worse experiences, especially in terms of customer service. Whether this comes to pass or not remains to be seen. However, where we have surveyed consumers about their experience with AI chatbots the majority are negative. It’s not that consumers object to AI per se. It’s that they want to have a choice and that choice has to involve being able to speak to a human being if needed. The future is blended: AI will play a role but it won’t replace the human touch.
If a retailer isn’t going to train and retain human beings, who are infinitely adaptable, can ask clarifying questions, and give practical advice based on firsthand use of the company’s products, then the retailer sure as heck isn’t going to invest even more into testing and refining technology that has nothing to do with its core business that would allow it to offer a superior customer experience. Rather, the retailer will go for the cheap option of a gussied-up predictive text phone tree, lay off staff, and tout this on its investor deck while giving bonuses to executives. We all know how this cycle goes; the consultants make good money and the whole exercise gets canned after 18 months.
Can it get worse? I think the mistrust comes from companies like Xfinity claiming they use ai to deliver the absolute worst customer service experience ever. Or some of the big banks. Or Facebook, that randomly puts people like me in “jail” because it’s programs are purely keyword driven with no sense of context or irony.
All one can think of is “ if this is ai I want nothing to do with it. Now, the truth is, there is no ai used anywhere in retail except in the supply chain, and that’s more machine learning than pure ai.
The hype is kicking those who proselytize the technology as the savior in the tush.
This. AI is not for customer service. I spend more time explaining what I want than explaining my needs to an actual person. i just don’t think people want cold and impersonal and won’t for a long time to come.
When digital customer service, fueled by AI, is done right, the experience can be better. The problem is some get it right and some don’t, so customers aren’t getting the consistent experience they want, which impacts their overall view of AI-supported customer service.
My annual customer service and CX research found that only 32% have successfully resolved a customer service issue using AI without human assistance.
There is no doubt that artificial intelligence is a monster from both heaven and hell. There is no way to tell at this point in time whether it will be more of one or the other. At present, AI is at the same stage as the internet was in the mid 1980s. While most of us recognized its high technology, its potential greatness, and its potential to change life as we knew it, our imaginations took us to both good and bad places.
For retailers or consumers, it is way too early at this stage to fully understand, appreciate, or despise features, benefits, and drawbacks, of which there might be too many, at least until the dust settles, if that ever occurs.
The initial effect of AI on customer service will be mostly positive, but as time progresses, there will be “issues” that will be hard to resolve. I think we can all agree the same has been true with the internet. Db
Let’s get one fact out of the way: many of the latest AI capabilities are not being built in-house by retailers, they are provided by third parties. That’s good news and it changes the conversation. The innovative and capable customer-service-enabling third-party solution providers that I’ve spoken to just over the last month have restored my faith in the AI-driven future. These companies have a vested interest in providing security, constant innovation/iteration, training, and service after the sale if only because they already face so much competition. In this case, reality is more powerful than perception.
Another reality is that these third party offerings are only as capable as the data which is provided to them by the retailer. Without well organized and managed data, the customer experience is suboptimal regardless of the software.
Absolutely true and thankfully, I hear that is getting better by the day for software that relies on retailer-supplied data. Some of the most exciting solutions grab data directly from shoppers. The ability to track, quantify, and act upon in-store activity is particularly exciting these days.
My hunch is that many consumers mistrust the potential for AI service bots based on their abysmal experiences interacting with dumb-as-a-rock chatbots and infuriating IVR phone systems. Both are programmed with human-designed algorithms that anticipate a very narrow set of generic service issues. As a result they usually thwart the delivery of quality service, not enable it.
As I understand things, a true artificial learning intelligence could likely be “trainable” to deliver better service experiences, but the training period will be probably be excruciating for humans who interact with it.
Until they can understand and reply helpfully to plain-language queries from naive customers, AI service bots will deliver more frustration than satisfaction.
Retailers who aspire to adopt AI customer service should think twice and test carefully. This may be one of those innovations where it is far better to be a “fast follower” than a “pioneer.”
When the AI-driven experience is all about forcing the customer to figure out how to get around the AI so they can get to someone who can actually help them, then consumers are definitely going to hate it. Not to mention the general cynicism that no savings will be invested in anything for the customer’s benefit or for the benefit of the employees who remain.
I know I find that I am reaching out to a brand because I have a problem that is more difficult to solve than can be answered by an FAQ – and by extension, a chatbot. Respect customers and give them an exit hatch. Or at least, analyze your logs to stop endless loops from happening.
One last thing: my uncle passed away in July, and I am helping my aunt close all of his accounts. You want to judge a company’s empathy for customers and the sophistication of their AI? Go through trying to tell a brand their customer is dead and see what happens next.
Amen sister, though I’m sorry about your uncle!
Well implemented AI should be a significant benefit for companies and customers. The “Well implemented” part is critical.
It’s still early days with few AI practitioners and little AI practice. Using 3rd parties with more and specific AI experience helps things get started, but AI skills need to be developed in house to ensure a high quality customer experience.
The wins for companies and customers look a lot alike. Companies can train the AI once and offer customers consistency across responses. No differences rep by rep to confound customers and result in more issues for companies. More AI agents can be made available to limit wait times for customers and improve the customer experience at nominal cost.
Expert human agents should always be an option for customers and the AI to select. But AI should be capable of solving a high percentage of the most common customer service issues quickly and accurately.
Periodic updates based on interactions keeps the AI “trained” up to the minute.
All of us have had infuriating experiences that leave us shouting “OPERATOR!!” into the phone.Let’s hope that the companies that implement thee technologies test them adequately, and provide customers with an alternative if necessary.
Yes. Consumers should be concerned about AI’s emerging presence in customer service as it is yet another downhill step in respect for the humanity of doing business. We are, after all, in a business of humans. The primary reasons given for AI are generally those of thriftiness — the eternal search for efficiency. Retailers must come to realize that an efficiency driven focus a race to the bottom — poor customer service, bad products, and more. The ONLY path to business health is one of effectiveness — bringing more people to the store, keeping them coming, and satisfying them (in general). That is a matter of product and experience. There is no indication AI helps a company take a step forward toward healthy companies.
For many years retailers have been using customer date to personalise offers. There has been a deal struck whereby customers share their personal data with retail brands they trust and in return retailers provide relevant and timely offers or positioning of products to customers.
AI is likely to add to the precision of these offers – from some respects this increases the convenience for the consumer because the offers or positionings become even more relevant.
AI also has a role to play with customer service – providing the ability for consumers to have their queries and issues fixed faster. Some may worry about ‘talking to a machine’ but if queries are fixed seamlessly this concern will be outweighed.
To give consumers peace of mind, retailers can always offer the ability to ‘talk to a human’ and indeed many of the uses of AI are focused on freeing employee time for them to spend more time with customers.
Consumers’ concerns about the use of AI for customer service are legitimate. Chatbots that have been leveraging AI for several years have been a disappointing and frustrating experience for savvy consumers. Personally, before I engage with a chatbot, I research the retailer’s website to answer my questions. I explore every option and by the time I resort to help, I probably know more than the chatbots. I have never had a problem solved by a chatbot and I just continue to request a “live agent” to get a human to solve the problem. I am still not convinced that AI is best suited for customer service, unless it is used to direct you to the right human.
The most promising areas for AI and generative AI are in improving operational efficiency, product recommendations and marketing.
It is too early to tell for sure what will happen. But consumer concerns should always worry retailers.
Customer facing AI in service to date has sucked, no doubt about it. But it’s because the technology has been immature and companies are over-indexing on “call deflection” as a key KPI. As LLMs become more powerful and human-like, service may be better than ever with genAI, just give it time.
AI for customer service most likely will follow a rules based system and it will be every bit as frustrating as listening to “for store address press one” and so on-worthless.
In some ways putting “real people” back on the phone is worse, because the business model dictates that cost trumps everything, and that means low wage workers following a script. Tedious and unhelpful.
AI is going to happen whether we want it to or not. The real issue is how much horrible customer service will consumers tolerate, and I believe the answer is a lot-primarily because this aspect of retail is a race to the bottom.