Photo: iStock | KanawatTH
Will Generative AI Improve Customers’ Experiences Interacting With Retailers?
Customer experience (CX) and customer service leaders gathered in San Diego this week for the Reuters Events: Customer Service & Experience West 2023 conference. The consensus was that customer care is a brand differentiator and strategic focus for companies–especially given that customer expectations, call volumes, service employee attrition and training costs are all up.
In a session titled “Generative AI in CX,” Shawn Strassburg, head of transformation, Best Buy, discussed how the consumer electronics retailer uses artificial intelligence to solve specific CX/contact center business problems.
“Three and a half years ago, where the company made a very on-purpose decision, that we were going to re-envision customer care with a key focus on customer obsession. And really, really understanding your customers’ words, their voice,” said Mr. Strassburg. The retailer is halfway through its transformation.
The company uses a top-down/bottom-up approach to understand customers using data to guide decision-making. Leadership, business, data and analytics, and finance teams secured a top-down commitment and alignment to truly understand customers and the internal customer care processes. Bottom-up insights come from front-line teams, call center agents’ experiences, and customers’ experiences interacting with Best Buy’s CX organization.
Best Buy created a lab staffed with top agents and discovered a disconnect between contact center system data and what was happening as customers called in. To improve agent performance and customer experience, Best Buy implemented AI for real-time audio transcription, automated contact summaries and contact intent detection.
Lab trials have delivered a 25 percent reduction in after-call work (ACW) as call agents no longer must manually compose and enter call history and actions. There was also an 18-second reduction in talk time.
“As we tested it out, we were anticipating about a 15-second reduction in just talk time, not a negative way, but in a super positive way because it was helping to build a foundation of trust and partnership within our agent base and in our customer base,” said Mr. Strassburg.
Best Buy is auditing its systems and processes to identify where to plug in AI capabilities to hear the voice of their customer better, discover the true intent of why they call for service and determine if it has succeeded in resolving issues.
Discussion Questions
DISCUSSION QUESTIONS: What role(s) do you see generative artificial intelligence playing for retailers looking to improve the customer experience? How does Best Buy’s approach compare to other retailers?
Poll
Loading …BrainTrust
Paula Rosenblum
Co-founder, RSR Research
Gene Detroyer
Professor, International Business, Guizhou University of Finance & Economics and University of Sanya, China.
You know, AI can support responses, but there is no substitute for knowledgeable staff. I know retailers are always looking for a way to save a buck, but neither has outsourcing customer service, creation of service bots nor will generative AI eliminate the need for people.
Sorry, this dog won’t hunt either. The more the industry obsesses on the term CX, the worse it seems to get. I think that’s because it’s CX-washing. Find me ONE customer who thinks service levels have improved anywhere. ONE.
Paula, I think the customer service levels have improved…with one BIG caveat! That is once you get to a real person.
I spent 20 minutes on the phone with an AI bot spelling and re-spelling my last name (many times) so I could get service. I may have to shorten my last name so I could get things done in the future.
Must add a caveat…the real person has to have a real command of the English language. All the scripts make me crazy
You coined a great term…CX-washing!
Paula – you’re missing the point of this implementation.
I applaud BBYs careful use and holistic approach to their implementation- they are not using a Bot to connect with customers but rather to automate procedures for agents (typing in call notes, etc) so they can better serve more customers.
In the long run, Generative AI will produce dramatic improvements and business value transformation. In the short run, I anticipate a lot of pretty clumsy applications and blunders. Things will get worse before they get better.
Well said. It’s all about employee enablement and evolution, not simply replacing people. Done well, everyone will be happy.
Good point. I see one gap in why generative AI can’t replace the support right now – the lack of digitized knowledge of the buyer’s behavior. I’m not talking about the knowledge base. I’m talking about the on-site or in-store customer journey and transferring it to the behavior model, so AI assistants can read the context of the customer’s request.
That’s when robots will grow in quality CX by an order of magnitude. I specifically work on that problem from a position of prompt modeling.
I love it when AI is talked about as an enabler of speed and efficiency rather than a simple replacement for human beings. Of course it will replace people along the way, just like all kinds of tools have done over the years. We can embrace and celebrate progress and evolution and still be wary of the potentially very messy side of what is about to unfold. Hats off to Best Buy for their focus on the customer.
Jeff, you are right on! In this case BBY is helping their agents be more efficient and I think this implementation will help them improve the overall experience of agents and customers.
It has nothing to do with customer experience – but that doesn’t mean it won’t help Best Buy. They may learn more about customers’ needs, although I would think if I were calling BB customer support it would be pretty obvious why I called.
Does anyone remember the guy who spoke at a retail conference, maybe 10-15 years ago, who demonstrated a conversation with AI that completely mimicked that of human interaction? He booked meetings, dinners and a hair cut completely using whatever software he was pitching. It never materialized.
That demonstration was seamless, it’s what AI in customer service will have to be or it will just be a more polished conversion still with less than satisfactory answers. Current technology is frustrating. For important service questions I still want to speak with a human.
May have been a Mechanical Turk
Maybe. It was supposed to be AI that was within reach.
This seems like a lot of explanation for the simple solution I seek. Simply, I want to tell the automated customer service agent my problem and have them solve it. I never imagine my problem is so unique that it is not listed in one of the six options they gave me.
To me, as they describe it, the best part is that the system will read my voice. Does that mean that my decibel level and choice of words will no longer morph into my expletives? Will the system understand when I am annoyed and deal with that first?
Over the years, call center interaction has improved tremendously. Even though it is clear that the last thing the retailer wants you to do is to talk to a person, when you finally get to a person, without exaggeration, they are terrific in dealing with the customer (me).
With a very helpful and brilliant associate, along with this technology, why would I not be thrilled?
Retailers should try to keep up with everything AI. It is already touching every aspect of retail. Yes, customer experience will always be key. That alone has multiple angles, needs, and inefficiencies where AI could play a role. Even the often awkward implementations of using bots for website chats can now be upgraded to GPT-powered interactions that can answer questions down to the SKU and product detail level within seconds. AI bots are already smarter than most realize, and getting better every minute.
My advice to anyone in retail, omnichannel, restaurant—anywhere that there’s overwhelming complexity—is to stop asking “if” questions and start asking “what if” and “how” questions about AI.
This is dead on. It isn’t a question of if this is going to happen. It IS happening. The assessment needs to be how to best incorporate AI tech into situations. As Ken said, “AI bots are already smarter than most realize, and getting better every minute.” True. Ignore at it your own risk.
I’m hardly a Luddite, but generative AI isn’t a magic wand. In fact it has definite Frankenstein’s Monster potential. These systems “learn” within the strict limits of their programming and sooner or later … and my bet is sooner … they will start responding in ways that may upset customers who are, in fact, already upset. We never seem to learn the overarching lesson of technological evolution. A screwdriver is a great tool. So is a hammer. But, it’s really hard to drive a nail with a screwdriver or use a hammer to unscrew something. Last year it was the Metaverse that was going to change 147% of the customer experience. This year its GAI. Can’t wait to see what Technological Salvation 2024 look like.
Generative AI is important to impliment today. It’s not expensive, and it serves a great purpose, which is the tier of support. It can answer basic questions that are asked repeatedly. What it cant do, as pointed out by our fellow BrainTrust’ers, is answer higher-level problems/questions and build relationships. But for basics, it can get customers answers quickly.
Best Buy’s approach works. It’s a process that must be built, tested, and constantly kept up-to-date. The company’s technology gurus are constantly auditing, tweaking, and making the necessary changes to deliver accurate and helpful information.
Exactly. Agreed!
No doubt there are a lot of Bad Bots out there and that dealing with them are super frustrating.
But BBY is not taking that approach. I’m impressed that they are plugging in proven AI tools in their process to help agents be more productive.
If you haven’t looked at some of the Apps (like otter.si) that record audio, writes notes, and generates summaries you’ll be surprised at how good they are: reduce manual effort in typing / note taking and allow one to concentrate on what’s being discussed
One of the key roles AI can play in helping retailers help their customers have a better experience is in returns mitigation and elimination. It’s a $800 billion+ problem and ripe for solving, so why not let AI take a swing? By better understanding why consumers return product AND why they keep product, merchants can design programs to minimize the former and accentuate the latter. Easy things like identifying shipping issues with in-store fulfillment and harder things like rewarding shoppers for not size bracketing all benefit from the use of AI and its ability to quickly spot patterns and offer actionable fixes.
Retailers need to be exceptionally careful. Implementing any generative AI systems at this point is a highly unstable endeavor as little is known about risks.
For example, this week I was reminded of how these systems can give nonsense answers because they have no understanding of the physical world in which customers live – they are merely word predictors.
In the ad world, it’s clear these systems give entirely mediocre wording to answers – like a middle schooler.
So retailers need to beware. And, remember that there is no early mover advantage for taking this highly unstable risk early. As Wiener’s laws observe about automation in airplanes, while it can be helpful many times, then things go wrong they go wrong in far worse ways than with human systems.
Generative AI is capturing a lot of attention right now and there’s no doubt it will play a growing role in helping customers have a more gratifying experience. However, in our opinion, the real exponential gains from AI will be more in operational areas – helping streamline and enrich the entire end-to-end process from when a shopper first sees an ad through to the purchase, fulfillment, delivery and support of the merchandise. Get that right and the experience takes care of itself.
Generative AI can help with the rudimentary things of the shopping experience, like wayfinding, brand insights, product details, deals, decision making. But the heavy lifting should remain with good people who are friendly, who are solutions oriented. If using AI can help both customers and employees with the boring stuff, then that frees up employees to focus on customer loyalty, and the qualitative part of shopping.
Customers expect an ear/sympathy and answers without delay. AI will replace employees when that is possible. Until then, customers are guinea pigs and it will not improve the customers’ experience.
DISCLOSURE – I work for Microsoft.
There are lots of potential uses for Generative AI in terms of customer experience. Initial thinking falls on things like more human-like chatbots but there are many other uses as well. Product development, customised/personalised marketing content, determining customer intent in order to improve overall personalisation recommendations are all things retailers can look at. It really is early days – winners are likely to learn by starting small then applying their learnings on larger initiatives.
In my opinion, Best Buy’s approach is very pragmatic. They are absolutely clear on which area they want to focus on and how they can create some value for their customers. We are at a very advanced stage of technology where generating audio transcripts, automated summaries, and data entry via AI is a possibility. Therefore, BBY’s automation efforts are definitely in the right direction and might as well motivate other retailers to take some cues.
Ultimately, human efforts may only be required where they are absolutely necessary. Otherwise, retailers would be happy to let the AI do all the “heavy lifting” and let their Customer Service Representatives (CSRs) focus their attention on resolving a customer’s issues.