3 Ways AI Explainability Will Help Retailers Better Understand Consumer Preferences

We are entering a new age of retail. Following a pandemic-induced boom, ecommerce is no longer the king of the retail industry, and brick-and-mortar shopping experiences are driving the industry’s growth as consumers hunger for experiences outside their own four walls. Customer preferences are evolving at a breakneck pace, and understanding their purchase behavior has never been more critical to success within the retail industry.

As consumers push retailers to merge the benefits of physical and digital shopping, technology will continue to become more critical to providing shopping experiences that keep customers coming back for more. The technology darling of the retail industry has long been artificial intelligence (AI). The best, brightest and biggest retailers have tapped into AI since its inception, evolving throughout the years to rely on AI to assist in mission-critical processes and customer-facing tasks.

Just look at Walmart in the present day — its AI technology is infused throughout their supply chain operations; it’s present in every digital search, and personalization efforts are used in AI-powered fitting room technology. Walmart’s latest conversational Text to Shop AI even allows customers to text what they need, with the AI ​​algorithm suggesting complementary products to customers.

But despite the benefits of AI, there is one major drawback in its use to create frictionless experiences for customers: its black box.

The “black box” problem at its core is a lack of AI’s ability to explain why it arrived at a certain decision or to expand on the factors that influenced a final outcome. For example, AI can easily make predictions around customer churn. However, without knowing why the algorithm is predicting customer churn it is impossible to take action to try to retain the customer.

The AI ​​systems that truly enable retailers to not just keep up with changing consumer preferences but to stay ahead of them are the systems that have explainability baked in, allowing leaders to look at the weighted factors driving the outcomes of any algorithm. By using AI explainability to better understand the consumer, retailers of any size can advance personalization, better retain customers, redefine the returns process and up the ante against competitors.

Here are some of the top ways AI explainability can not only create enhanced, more frictionless experiences for consumers but can also help retailers better understand their customers in the long run.

  1. Personalized predictions you can understand: AI explainability, or the ability to analyze what factors lead to an algorithm’s decision, is critical in understanding buying behavior and in turn producing more accurate product recommendations via online shopping or in email promotions.

    According to a recent InRule survey of 1,000 US-based consumers, 20% reported that they’ve never bought a product from an online recommendation. Furthermore, with 43% of respondents reporting that they sometimes purchase a product recommended by a retailer during the shopping experience, it is evident that consumers will buy if the right product is shown. The key is to ensure that AI/ML recommendation models are clued into organizational data and can outline for human-in-the-loop knowledge workers exactly what the customer is interested in, and why, to ensure continuous improvement in models that build customer loyalty .

  2. Retain your customers: Brands need to better understand why customers have a certain preference on the type of products they’re purchasing, and ensure this data is fed back into the training of the model.

    Machine learning can easily make predictions around customer churn — there’s a huge difference between an algorithm stating that there’s a 90% chance this customer will not revisit the website for another year versus that there’s a 90% chance the customer will not revisit the website unless they receive 20% off their order. In this instance, the algorithm is not only predicting customer churn, but it’s using data to inform the retailer of extreme price sensitivity and offering a solution to drive them back to the website sooner than would happen naturally.

  3. End the guesswork of returns: For consumers and retailers, a return presents backend challenges and roadblocks. After all, 37% of consumers reported that they expect their return refund to be processed within 48 hours, yet the industry standard refund time frame is three to five days. Additionally, 26% of consumers stated that they are deterred from online shopping if they had complications with a previous return process.

    While retailers may have figured out better logistical processes for physical items, backend processing can be streamlined with advanced AI/ML-backed automation that expedites returns, explains why a customer made a return and provides recommendations on how to retain a customer despite dissatisfaction with a product.

AI explainability will not only help retailers understand customers better, but it can make recommendations on how to act when customers are at risk of stepping back from the brand. Coupled together, retailers that use this technology can create standout loyalty programs and promotional materials. While AI is currently being used in personalization, retention and returns efforts, AI explainability is the competitive differentiator for retailers to trump competitors with newfound sophistication and understanding of the customer.


Rik Chomko co-founded InRule Technology with Loren Goodman in 2002. He became CEO in 2015 after serving as COO since 2012. Chomko also served as Chief Product Officer prior to his role as COO. Before co-founding InRule, Chomko was CTO with Calypso Systems, a Chicago-based consulting firm. He also worked for Health Care Service Corporation from 1991 to 1995. With more than 25 years of experience innovating and implementing technology-based solutions, Chomko has focused his efforts on ROI-driven digital transformation and automation of enterprise systems and processes. While in product-focused roles, he led the development and enhancement of InRule’s most successful offerings, including integrations with .NET, Java, JavaScript, Dynamics 365 and Salesforce. Under his leadership, InRule is recognized by top analyst firm Forrester as the leading business rule management system for the .NET platform.

Leave a Comment

Your email address will not be published.