The New AI-Powered Commerce

Abstract colorful  illustrations of a colorful phone and Ipad surrounded by colorful shapes.

Artificial intelligence (AI) has inspired the imagination since Alan Turing coined the term “thinking machines” in the 1950s. Countless books and movies have explored how AI might shape our future while data scientists developed ever-more sophisticated algorithms. By 1997, IBM’s AI, nicknamed Deep Blue, was able to handily defeat chess champion Gary Kasparov.

Today, the AI-powered future is here — and we’re living it. AI now affects how we interact with the world around us on a daily basis, and it’s already a commercial powerhouse. According to a recent survey by Adobe and Forrester, 56 percent of businesses are investing in AI. And they are using it to bring efficiencies across their entire organizations, from customer service to operations workflows.

E-commerce represents one of the most promising applications of AI. For digital businesses, it has the potential to improve the shopping experience, simplify and streamline digital merchandising, and make fulfillment smarter and more efficient. And, rather than replace human marketers, it is more likely to free them from repetitive tasks. According to a Deloitte survey of CMOs, a mere 1.7 percent said new technologies are replacing marketing employees in their companies by “a great deal.”

This post explores three powerful ways AI is transforming how consumers and businesses shop online. It includes AI-powered e-commerce features that are available today — and some that may be coming soon to a future near you.

Improving the digital customer experience

Research suggests what everyone intuitively knows — companies that deliver a better customer experience are more financially successful. In a study by Adobe and Econsultancy, customer experience leaders typically outpaced their sectors — and were three times more likely to have “significantly outpaced” their sectors than mainstream organizations.

Investments in improving the customer experience, then, are likely to deliver substantial returns — and companies are betting big on AI. According to research by McKinsey, “high AI performers invest more of their digital budgets in AI than their counterparts and are more likely to increase their AI investments in the next three years.”

More accurate search

One of the more frustrating experiences when shopping online is when you are looking for a product and cannot simply find it. You may type in many different search terms, in multiple combinations, and still come up dry. This is because search on many e-commerce platforms is powered by manual tags, which often correspond to unique and unintuitive product signifiers developed by operations or engineering teams.

AI can automate and simplify this process. Using natural language processing, it analyzes the language customers use to search for products and learns which “real world” words and phrases are synonymous with your product tags. For example, let’s say you sell a product tagged in your catalog as a “sweetened carbonated beverage.” AI can help surface this product in customer search results for “soda,” “pop,” and other terms. Also, AI can update product tags with their natural language equivalents, making search even faster for customers — and more efficient for your marketing team.

Personalized product recommendations

Personalized product recommendations are another way AI can enhance the customer experience. In fact, these kind of custom-tailored recommendations are already an entrenched part of today’s online shopping experience. Anyone who shops on Amazon will see the home page covered with personalized recommendations based on their browsing and shopping history. And Netflix viewers are familiar with their “binge-worthy” suggestions.

Both Amazon and Netflix use costly, proprietary AI to power their recommendations engines. And, until recently, personalized product recommendations were only a realistic option for businesses with big IT teams and technology budgets. However, that is changing fast. AI has become much more accessible through purpose-built extensions and even e-commerce platforms.

Moreover, personalized product recommendations are increasingly becoming a requirement for both business buyers and consumers. Seventy percent of business buyers say personalized recommendations help them to obtain more value from their vendors. And 47 percent of consumers check Amazon if the brand they are shopping with doesn’t provide product suggestions that are relevant. All this suggests that if you don’t offer personalized product recommendations, you risk losing business.

Even more CX applications for AI

Of course, intelligent search and personalized product recommendations are just two examples of how AI can significantly improve the customer experience. It can also predict the kinds of content and offers shoppers most want to see, guide buyers through complicated purchase processes, and even have “human-like” conversations via chat.

Simplifying and streamlining merchandising

The e-commerce experience does not exist in a vacuum. It relies on merchandising that shows customers how products look, feel, and function. And merchandising can be surprisingly complex and laborious to maintain. As seasons change and trends evolve, merchandising must be continually updated — and many online sellers still do it manually. This can mean editing product descriptions on multiple pages, uploading and tagging new images, posting new promotions, etc.

In addition to keeping merchandising up-to-date, e-commerce businesses need to understand how effective it is. But existing technology makes it very difficult to analyze how different product descriptions, images, and related content contribute to conversions. Also, even creating the necessary product imagery, content, and video can be expensive and time-consuming, and require engaging multiple creative teams or external agencies.

How AI could help

AI promises to make merchandising both easier to develop and maintain and more effective. For example, AI could automatically search a product information database for changes and make updates to the e-commerce system, no human intervention required. Similarly, AI could track how different product images and copy perform in real time and continually make adjustments to optimize your conversion rates. It could do the same for content and offers.

Another way AI could support merchandising is to automate content creation. For example, it could automatically combine modular content (copy, images, video clips, etc.) to create new custom content. In addition, AI is capable of building personalized content on the fly for individual customers who visit your website. AI can already write copy. It could generate new imagery and even shoot video via drone.

While this kind of automated merchandising is not readily available today without extensive custom development, I believe at least some of this functionality will become accessible over the next two-to-three years.

Delivering intelligent fulfillment

Another essential part of the overall e-commerce experience is fulfillment. This refers to finding products in your warehouse and getting them wherever they need to go. As you add new products, sell higher volumes, and serve new locations, fulfillment can become extraordinarily complex. It’s even harder when you have both digital and brick-and-mortar stores — and inventory stored in multiple locations. Mistakes — like failing to ship products from the closest store or warehouse to your customer — can be expensive.

Moreover, effective fulfillment requires managing data flows between multiple systems, including your e-commerce platform, ERP, OMS, POS, CRM, and more. Many businesses still rely on batch processing to get data where it needs to go. Unfortunately, periodically run batch processes can lead to disconnects in which your e-commerce listing can say a product is available, but it is really on backorder.

How AI could help

AI could make fulfillment smarter and more efficient. For example, it could optimize fulfillment for delivery times and costs for every order. With machine learning, it could track trends — e.g., late deliveries — and automatically adjust strategies to improve performance. It could also use weather and other data to predict fulfillment issues and avoid them by not shipping from warehouses in affected areas — or even changing the shipping options available for certain product listings.

Again, while this level of AI-powered fulfillment automation is not yet available today, the building blocks are already here. It could realistically be available within the next two to three years.

See what’s next

AI promises to have a profound effect on all aspects of e-commerce, and the pace of change is accelerating. Already, customers expect some forms of AI-powered commerce, like personalized product recommendations and offers. Businesses of all kinds are making big bets on AI — and those that don’t risk being left behind. To stay ahead of the curve, I recommend prioritizing digital transformation and asking your e-commerce platform provider about their roadmap for AI.

Watch the Adobe Summit session on how AI can drive results.