Home Depot and Lowe’s launched AI assistants in the same week last year, and since then, the retailers have each rolled out fresh capabilities and features — often within weeks of each other.
Home Depot’s Magic Apron and Lowe’s Mylow share a goal in helping customers navigate the complexities of a home improvement journey but take different avenues to get there.
Home Depot’s entire CX strategy is grounded in solving customer problems, according to Sarah Post, VP of interconnected experience at Home Depot. Magic Apron and the retailer’s other AI investments are focused on finding areas where it can help customers make progress in their projects regardless of how they shop.
“Our customers don't necessarily care or think about digital or in-store, or what is AI or what's not,” Post told CX Dive. “They care about two different things: One, help me solve my problems, and two, help my dreams come true. Those are the two core use cases that we solve around.”
Lowe’s sees every customer journey as a mission, whether that’s a few simple supplies or everything for a massive remodeling process, according to Neelima Sharma, SVP of omnichannel and e-commerce technology at Lowe’s. Mylow was designed to support both paths.
“We're supporting everyday purchases by knowing the context and helping them solve the problem very quickly in a frictionless way,” Sharma told CX Dive.
While the goal of supporting customers is shared between the retailers, recent updates for Magic Apron and Mylow showcase how each is taking its own approach.
In August, Home Depot added localized store knowledge to Magic Apron to help it offer better in-store guidance. That same month, Lowe’s announced Mylow Memory, which enhances its AI assistant with customer context.
Making full use of AI-powered personalization comes down to understanding customer intent on a deep level, according to Nikki Baird, VP of strategy and product at Aptos Retail. Retailers have long had access to the signals, but as third-party AI platforms become more popular, retailers will have to adapt to maintain their connection with the customer.
Magic Apron offers support from start to finish
Magic Apron puts an emphasis on supporting the entire customer journey, no matter where it starts or ends.
The AI assistant’s capabilities are built around three common customer needs, according to Post. The first is that shoppers are constantly spending time searching for specific items. That led to the addition of local store information that can help customers locate specific products and tailor its responses to questions based on what’s available at their local Home Depot store.
The second need is answering questions about how to use products or tools, which the AI assistant can help with through tutorials. The third need is walking customers through their projects, and Magic Apron aims to make answering questions simpler through a combination of visual search and voice tools.
“It's multimodal,” Post said. “A lot of different use cases that our customers come into the store and need help with are now accelerated and augmented with our Magic Apron experience.”
Customer engagement with Magic Apron is strong and growing, with the tool receiving millions of questions every month, executives said on an August earnings call.
Customers often start their journeys in need of inspiration and will browse the website or ask Magic Apron a question about what they need to achieve, according to Post. The multimodal nature of the AI assistant means customers can use voice, text or a photo to start.
“The visualization aspect of Magic Apron helps customers understand exactly where and what that product is and where you can find it in the store,” Post said. “It happens all the time.”
Tools that help streamline the shopping experience by making product and knowledge discovery easier are a key use case for AI, according to Garry Heon, SVP and sector head of retail and consumer goods, communications, media and entertainment at QualityAI.
He likes to handle his home improvement projects personally and has found that Magic Apron has helped save time.
“When customers go to the store, they're not wandering around for an hour,” Heon said. “They may find associates and ask for help, but they're going to have targeted questions because they're informed by the AI.”
Mylow thrives on customer context
Lowe’s wants Mylow to feel like an associate who knows you personally, and Mylow Memory is designed to bring the associate one step closer to that reality.
Mylow Memory builds on the assistant’s existing personalization capabilities by giving it additional context, according to Sharma. This helps the tool offer advice beyond a generic AI answer that could work for anyone.
“The whole idea with Mylow Memory is that every single time you interact with us, it's like that super associate who has home improvement expertise but also has context about you,” Sharma said.
Achieving this level of support requires understanding not just the customer, but the context of their project, according to Sharma. She compares it to a friend.
“Think of a friend that you've had for a long time,” Sharma said. “You’ve shared experiences. You've shared information with each other. They understand you. So when you ask them for advice, you have a lot of trust that they'll give you good advice because they really understand you and your situation.”
LLM technology has made bringing conversational interactions to life much easier than they were in the past, Sharma said.
LLMs have come with their own challenge in the form of customer switching from search to third-party platforms like ChatGPT, according to Baird. Research has shown that customers trust retailers’ own assistants over third-party AI tools, but retailers need customer insights as fuel for first-party AI.
“We've built a whole ecosystem around divining customer intent from the search terms that they use, from the questions that they ask, from the clicks that they make when they're on the site, and I think a lot of that is going dark and getting shifted to other places,” Baird told CX Dive. “It's going to be more important for retailers to have more insight into intent.”
Lowe’s has its own ecosystem to draw on, according to Sharma. Mylow Memory uses data including what products customers have browsed, what they have asked about in the past and what projects they have already completed with Lowe’s help to provide context.
Mylow has proven successful thus far, President and CEO Marvin Ellison said at a conference last month. The assistant has fielded more than 25 million questions from shoppers and associates, and the conversion rate of customers who use Mylow online is three times higher than those that don’t.
“Any time conversion is three times better, that's a tangible benefit that we can connect directly to the fact that this AI assistant is giving customers some information that they need,” Ellison said during the conference.
Mylow and Magic Apron are just getting started
Both Lowe’s and Home Depot are designing their AI assistants around rapidly changing journeys, and the retailers’ work is far from done.
Consumer habits are already evolving, and the impact will only grow as general purpose AI gets better at understanding its users, according to Sharma. She expects customers to become more reliant on the technology over time, the same way search engines have become an everyday tool.
“The technology is so nascent, and the use and the adoption is so rapid,” Sharma said. “It's how customers are looking for information, shopping for products, getting recommendations on products — it's already changed so significantly and so fast.”
This will be particularly important for home improvement projects, according to Sharma. Most consumers won’t remodel their kitchen more than once or twice in a lifetime, but an AI assistant can serve as an expert adviser that understands the customer and the project.
Home improvement is “a really hard category,” Post said. Customers need support that offers reliable expertise as it helps them solve their problems.
The retailer has learned from the prompts customers use when they interact with Magic Apron, according to Post. This information is helping the company sharpen the AI to support customers, and it will continue to study how customers want to interact with AI assistants as technology and expectations evolve.
“The beauty of AI is that it's this natural human language,” Post said. “You don't have to necessarily navigate someone's interface. You can actually have a conversation with it, but we'll constantly be experimenting on how to better solve our customers' problems with whatever interface comes about.”
