How AI plays into every step in the retail customer journey
When it comes to misconceptions about artificial intelligence, the likely top one is that retailers tend to view AI too narrowly and don't realize how the technology can help solve every retail problem.
Simply, AI isn't just about weaving chat bots into the retail customer experience or just streamlining operations. It's about teaching retailers about their customers via deep learning and genetic algorithms, providing insight on how users interact with products, how messages resonate with shoppers and how to better market products.
"Every step of the customer journey will be transformed by retail, from the ads and content that bring in customers to their experience on the site to logistics and shipping and even product design. AI won't be replacing marketers or site architects, but it will be making them far more powerful," explained Andy Narayanan, vice president of intelligence commerce at Sentient, in an email interview with Retail Customer Experience.
For example, most retail sites house some sort of recommendation engine to help shoppers find products that fit their sense of style. Most however, said Narayanan, are still "coarse" at this point and tend to lump shoppers into groups.
"They have trouble digging deep into catalogs to find items that might be a bit less popular but fit a user's desires and intent perfectly," he said. "And they take a lot of time and data to set up. AI can solve for these exact issues. It can learn, in-the-moment, what a shopper likes just by analyzing images.
"No more product tagging and historical data: this is recommendations based on real-time behavior leveraging an asset that retailers already have — namely, their product images. This means a shopper can find products they love fast. And, more importantly, it means systems that treat individual shoppers individually."
As women's wear and garment retailer Cosabella is learning it can play a pivotal role in driving sales and conversions as well as attaining deeper insight on customer wants, needs and interests.
Cosabella delved into AI after experiencing a period of flat-lining with a digital ad agency.
"This forced us to look into other opportunities," Courtney Connelly, Cosabella's marketing director, said in an email interview with Retail Customer Experience. "We considered three options: find a new agency, build a larger in-house team or explore the AI world. We obviously explored the AI space, thinking it would be cost prohibitive, but in fact, found some pretty incredible and affordable options. We were looking for programs to help us scale quickly."
The family-owned retailer, known for its innovative lingerie designs, has been working with Sentient for nearly eight months and chose Sentient given its extensive background in AI. It launched various AI programs with Sentient in the last quarter of 2016 and is currently running three AI-based programs.
"We love the automation and time-savings components, especially in the automatic mutation of candidates in our experiments," said Connelly.
Cosabella's first AI experience was geared toward boosting revenue via conversations. It tested button colors, marketing messages and image sizes. A 30-day experiment brought a 30 percent lift in conversions, shared Connelly.
"Our most recent experiment focused on the newsletter subscription window. We had already doubled our subscribers but found the homepage pop-up was contributing a very small percent. The experiment played with size, color, timing, fonts, messaging and placement and produced a winning candidate that increased subscriptions by 2000 percent," she said.
Cosabella's AI experience reflects why more retailers embracing AI, said Narayanan.
"We are at the point where more and more retailers are waking up to the fact that if they don't implement AI, they may get left behind. The great news is that when we start talking to retailers they begin to understand that AI is not just some fad that you market to your customers, it is something that you can implement across your entire customer journey, from first click to checkout you can immediately see a difference in engagement and conversions," he said.
One big AI return-on-investment revolves around shoppable editorial images, added Narayanan.
"Imagine seeing an outfit you like on Pinterest or Facebook and then clicking on it and being taken to a page on a retailer's store front where you can personalize that ‘look' to your individual style. This is when retailer's see the true power of AI and how it can directly impact their bottom line," he said.
Adopting AI does bring some lessons, said Connelly, who recommends retailers take the time to understand how AI works. She acknowledged that was her biggest hurdle.
"A lot of times it is instinctive for humans to put in parameters that could in fact contaminate or stifle the AI's processes. I could have saved myself a lot of frustration if I would have taken more time in digesting how the programs work," she said.
"AI requires freedom and options, much like a very intelligent employee. This is very different from how we tend to think about programs, we tend to see programs as pure tools in our hands. The best way to sum it up is that, if you have a very talented person on your team, you will get the most out of that person by giving them some freedom and not micro-managing them. The same applies to AI in many ways."
Narayanan believes AI adoption will be fast and furious in the next few years. A recent Gartner research prediction states 85 percent of retail customer interaction will be managed by AI by the year 2020. That, said Narayanan, is a bit on the low end given how AI is providing greater accuracy in predicting a consumer's style at the exact moment they're shopping. Such exact predictions will lead to fewer returned items and better stocking ability — which equates to happier customers and big savings for retailers.
"Personally, I think it will but even more pervasive than that. We will get to a point where every retail purchase you make online will have been influenced by AI in one way or another. Buying will become so frictionless for shoppers that they won't even know that AI is helping them. Every retail problem will have been solved by AI," Narayanan said.
Topics: Assisted Selling, Automated Retail / Vending, Consumer Behavior, CRM, Customer Experience, Customer Service, eCommerce, Marketing, Merchandising, Omnichannel / Multichannel, Online Retailing, Retail - Analytics, Retail - General, Shopper Marketing, Technology
Judy Mottl / Judy Mottl is an experienced editor, reporter and blogger who has worked for top media including AOL, InformationWeek and InternetNews. She’s written everything from breaking news to in-depth trends. She loves a great pitch so email here, follow on Twitter and connect on LinkedIn.