Personalization at scale for retailers: Artificial intelligence, authentic retail personalization

For as long as retail has been around, personalization has been at the heart of it. The ability for a store associate to suggest the perfect accessory for an outfit, give an informed opinion on which paint color will go best with the sofa fabric, recommend the perfect wine to compliment your meal kit — this kind of empathetic human-to-human interaction has always been the definition of outstanding retail experiences. It’s how retailers stood out and why customers came back.

But it’s a different retail world now. While face-to-face encounters still matter, interactions between the retailer and customer happen in countless places, both physical and digital. As if that weren’t enough, customer expectations in this new retail world are rapidly expanding. Dialed-in recommendations, the ability to buy online and pick up in-store, omnichannel inventory, on-demand and highly-tailored customer service — the list goes on.

In other words, it’s the same personal touch customers have always expected. Except now they want it wherever they are and however they interact with a retailer, with understanding of their preferences and desires at that very moment. Fortunately, machine learning and artificial intelligence capabilities are expanding right along with customer expectations, dramatically increasing the retailer’s ability to deliver exceptional personalization at scale.

Let’s start by getting something straight.

Although the terms are often used interchangeably, machine learning and AI aren’t the same thing. Machine learning uses algorithms that learn from data without explicit rules. These algorithms build models to make data-driven predictions or decisions rather than strictly follow instructions. AI, on the other hand, performs tasks that would usually require human intelligence, such as image recognition, language processing, and complex social interactions.

But machine learning and AI share much in common, and both are increasingly essential for achieving the level of personalization retailers must provide to meet basic customer expectations — let alone exceed them.

Moving up to machine learning and AI.

As discussed in an earlier article, it’s possible to personalize experiences using rules — no machine learning or AI required. But rules-based personalization has its limits, since it typically only considers a single data point — such as new visitor or return visitor, or approximate geo-location based on IP address.

Machine learning is more sophisticated than that. It can analyze enormous and seemingly disconnected sets of data deeply and quickly — and then act on that analysis. That’s why retailers are replacing rule-based approaches on their web and app experiences with predictive experiences driven by machine learning, accurately shaped by the products individuals actually want.

And it’s not just the huge ecommerce companies that are turning to machine learning and AI for personalization at scale. According to a global survey conducted by Forrester Consulting and commissioned by Emarsys, 54 percent of retail marketers are using AI-driven personalization across channels to drive growth in their business. A majority are also using it to understand customer behavior and manage real-time customer interactions across channels and touchpoints.

Ways retail marketers are using AI to drive innovation and growth.

Source: Emarsys and Forrester

Possibilities powered by AI and machine learning.

It’s easy to be jaded by all the talk about artificial intelligence and machine learning, but its powers are real — especially when it comes to personalization at scale for retailers. Here are just a few examples:

All data, no human bias.

AI and machine learning can draw on data from every source imaginable — supply chain to CRM, online to in-store, behavioral to contextual. It can then analyze these seemingly disconnected sets of data deeply and quickly to drive personalization at scale. It also removes human bias — incorrect assumptions even savvy retail marketers have about their audiences and what impacts them — and can spot opportunities for increased conversions and revenue that even the best retailers miss.

Fluid experiences.

With AI and machine learning, retail experiences can be automatically adjusted based on situational context. Depending on where a customer is in the journey or even inside a store, messages and content will adapt accordingly. This can add an entirely new, highly personalized dimension to the retail customer experience.

Shoppers to advocates.
Retailers are using AI and machine learning to automatically identify the high-value customers who generate the vast majority of revenue, then use it to deliver high-performance loyalty experiences. At the same time, retailers can efficiently and effectively engage lower-value customers, plus identify who’s most likely to become high value over time.

Impact beyond personal screens.

From voice interfaces to in-store interactivity, the role of machine learning and AI in retail extends beyond phones and laptops. For example, this AI-powered supermarket Amazon Go brings the “frictionless experience” model that’s so central to online experiences to the physical world. Amazon Go opened to the public in early 2018, and it’s already looking to expand into new markets. Others, like Albertsons, are also exploring the possibilities. As AI capabilities grow and become increasingly affordable, the frictionless in-store experience will likely become yet another standard customer expectation.

Adobe: AI smart.

For us, machine learning and artificial intelligence aren’t new. Adobe Target has used both for more than a decade. As a result, it offers substantial personalization at scale for brands all over the world. For example, it uses machine learning and AI to deliver personalized content based on real-time data. Retailers are also using Target for everything from personalized recommendations to automated offers.

But it’s not just a Target thing. Machine learning and AI are integral across all Adobe offerings — Creative Cloud, Document Cloud, and Adobe Experience Cloud. Adobe Sensei, our artificial intelligence and machine learning engine, is helping retailers discover hidden opportunities and accelerate delivering relevant experiences to every customer. And its capabilities are growing regularly, exiting beta and entering retailers’ day-to-day customer experience creation at scale.

AI of the not-distant future.

This is just a start. To create the next generation of AI, Adobe R&D is using ensembles of sophisticated algorithms and pulling from relevant research in areas such as natural language processing. Some of this work will likely have big impact on retail brands and their customers.

As an example, the R&D team is looking for ways to decode how those ensembles of numerous complex algorithms come to the conclusions they do. This might help marketers learn from AI decisions. AI’s approach to chess and Go has influenced how top international champions in those games play, and it similarly might impact how retailers approach building customer experiences.

Next up — meaningful creative and content.

Personalization at scale dramatically increases creative and content volume, but it doesn’t decrease customer demand for quality. Fortunately, AI and machine learning can help here too. The next article in this series will get specific on how to achieve the volume of content and creative necessary for personalization at scale.

“Adobe Transforms Personalization with Artificial Intelligence,” Adobe Newsroom, July 27, 2017.

“Albertsons to Pilot Amazon Go-Style Technology in Stores, at Gas Pumps,” Progressive Grocer, May 21, 2018.

“Building Trust and Confidence: AI Marketing Readiness in Retail and eCommerce,” Forrester, commissioned by Emarsys, July 2017.

Krista Garcia, “Shoppers Say They’re Self-Sufficient,” eMarketer Retail, March 18, 2018.

“Reinventing Retail: 10th Annual Zebra Shopper Study,” Zebra Technologies, November 2017.

Taylor Soper, “Amazon Go to open in Chicago and San Francisco, expanding cashier-less store footprint,” GeekWire, May 14, 2018.