AI in retail delivers next-generation online personalization

Today, many companies market to broad segments because it's more efficient than one-to-one outreach. But each consumer has specific needs and wants. AI promises to fix this.

Online retail has been around for more than 20 years -- artificial intelligence, even longer. The combination of the two, coupled with current computing power and data growth, can offer today's retailers a personalization punch that lifts sales.

According to "The State of Retailing Online 2017," a survey of 74 retailers conducted by the National Retail Federation and Forrester Research, the median conversion rate for retailers in 2016 was 3%. It's easy to see how AI in retail can lead to a small change in conversion rates.

There are some aspects of AI that must be addressed before an organization can fully embrace the results of artificial intelligence, however, including the following:

Recommendations, not personalization: Many online sites feature static, quasi-personalized recommendations, such as people who looked at this product actually purchased that product listings, or results that link to similar products or the top products in a given category. While recommendations can surface products that consumers may be interested in, they can also disrupt the shopping experience by forcing consumers to wade through products of interest to others.

Recommendations are very common, but fall short because they're not always relevant and can require too much work on the part of the consumer.

Manual merchandising and website optimization: Many online retailers wade through an avalanche of data to improve conversion rates with better merchandising, testing and website optimization. But this is limited by the web teams' ability to harness vast amounts of data, respond rapidly to change and treat each shopper as an individual. With more and more data coming in and an accelerating pace of change, an approach limited by people is most likely doomed to fail.

Benefits of using artificial intelligence in retail

That's why more and more companies are including AI in their retail strategies to create dynamic, personalized experiences tailored to each customer. In an ideal world, this personalization is also invisible -- it's just a part of the experience; so seamless that consumers don't even know it's there. This is not about hiding personalization, but rather integrating it deep into the customer experience.

Components of AI
Chatbots are just one type of artificial technology available to organizations. Here's a look at the components of AI offerings.

The end goal is to help people find what they want, suggest things they're likely to buy even if they don't know they want them, and up-sell and cross-sell. All of this must also be done with as little friction -- confusion, extra steps, wasted time on things customers don't want -- as possible.

In this new world of AI in retail, the entire experience is personalized to the individual user based on myriad internal and external data points. Buying signals, such as the shopper's viewing and buying history, social media behaviors, and demographics, are also factored into the customer journey.

So how does AI change the online retail experience? The following are four examples:

  1. AI incorporates hundreds of signals to more accurately determine intent and context. AI can continually ingest and utilize hundreds of internal and external buying signals that have been positively correlated with, for example, a higher propensity to purchase, higher shopping cart value or higher customer lifetime value. AI can also incorporate behavioral data about a shopper -- pages visited, previous purchases, time on the site, mouse movement -- and context to explain the user's interests and intent.
  2. AI can help develop fully personalized experiences correlated to purchase behavior. AI can use buying signals to drive a user experience that also correlates with an increased likelihood to add to a shopping cart, higher shopping cart value, etc. The personalized experience might include placement, size and inclusion of everything from product selection -- which products are shown to the shopper -- to copy, images and call-to-action buttons, such as Add to Shopping Cart. AI in retail technology can even help personalize site views: If Bob responds better to red buttons, but Becky responds better to green buttons, each would see buttons in the color to which each is most likely to respond best.
  3. AI provides opportunities for rapid learning for rapid change. In addition to faster processing, AI also enables continual improvement, as machine learning learns from experience, automatically testing and reusing the personalization and tactics most likely to lead to higher value sales, while dropping less successful tactics. These tests take place continually and are not limited to a human's ability to arrange and monitor tests.
  4. AI for retail settings enables background personalization. As mentioned earlier, this personalization drives an experience that is so immersive that it feels like just part of the process, not an added layer to be waded through.

Considerations when rolling out AI for retail

While AI-driven personalization has a lot of potential, some caution should be taken.

  • AI requires flexibility: In order for all aspects of the buyer experience to be customized, you will need a platform that enables that level of customization, such as flexible layouts that can vary by shopper, variable color schemes and more. The retailer brand fades in this environment, becoming less about a uniform color and look and feel, and more about an experience that is highly tailored to the individual shopper.
  • AI in retail can lead to increased content demands: More variables may mean increased content demands for different types of copy, category and product descriptions, promotions, product images, and more. At the same time, AI will come to generate more and more of that content directly.
  • Continuous adjustments are needed: Obviously, the fast pace of change requires even more focus on reviewing results to ensure that goals are being met, as well as having the tools in place for business people to continually make the required adjustments.
  • IT staff members may need to fine-tune their skills: Although it may not seem that way at first, AI is not actually about ending employees' roles in online retailing. Rather, as AI takes on the repetitive data crunching tasks, people will need to focus more and more on analytical and creative skills.

Next Steps

Discover how BOPUS is changing the customer experience.

Dig Deeper on Customer analytics