Walmart develops AI tools to help suppliers better understand customer data

The new AI features include a conversational tool to help suppliers better understand Scintilla’s data, another tool offering summaries of customer survey results, and enhanced marketing and advertising strategy recommendations.

The conversational tool, available early next year, will help assist users on what specific metrics mean, how to use them in their business and how they are calculated. Scintilla tracks metrics, such as sales, category performance, customer penetration and web traffic, from Walmart’s stores and e-commerce platforms.

“That tends to always be the first point of education: getting people to actually have a single version of truth, including the metrics they’re reading themselves and how they define it,” said Mark Hardy, head of Walmart Data Ventures. “This is helping users become more sophisticated themselves, in terms of the application of the data to the business decisions they have to make and making sure that there is one consistent way of applying it.”

Hardy expects suppliers to benefit the most from this in understanding data on customer-related metrics. For decades, he said, Walmart and its suppliers have used operational data to forecast and make decisions about where to put products and inventory levels. Scintilla, however, adds that there are many new data sets relating to customer behavior and perception that suppliers may not yet be familiar with.

“We need to always start with a customer, not start with our sales and operational metrics,” Hardy said. “Being able to educate people at scale on customer metrics, customer behaviors, customer attitudes — that, I think, is where this will become really impactful.”

The AI-generated research summaries, expected to be available by the end of this year, are intended to help suppliers understand survey insights. Through Scintilla, brands can run surveys with customers who are part of the Walmart Customer Spark Community, an invitation-only program that offers customers gift cards in exchange for participating and has expanded to in-home product tests this year.

Today, when the company does surveys, either a team of researchers at Walmart analyzes the survey data or suppliers can do this themselves. The information can help brands decide how they should improve assortment choices, direct campaign messaging or name products.

The company plans to make AI-powered research summaries of surveys available right at the end of the studies. Hardy said that by using AI to generate conclusions and inferences immediately after studies, suppliers can spend more time thinking about their strategies instead of creating graphs and charts on their own.

“What we’re doing by using AI is really cutting out all that time and labor required to do that last step,” Hardy said.

Hardy said the company is also working on a collaboration tool for suppliers to collaborate with researchers as they develop surveys for Walmart shoppers.

Finally, the company plans to incorporate AI into Insights Activation, a module within Scintilla that launched last year. It extracts Scintilla data to use for Walmart display-ad campaigns through the Walmart Connect retail media network. In May, Scintilla added new custom-audience targeting tools that can help brands target ads and other marketing efforts based on price sensitivity and demographics, like gender and age.

“It’s constantly looking at your customers’ shopping behaviors and looking for changes in those behaviors, and putting recommendations on how you can align some of the behaviors that we’re seeing with strategies around marketing,” Hardy said.

Andrew Lipsman of Media, Ads + Commerce, a retail media analyst and consultant, said within retail media, customer insights or audience data form the basis for activations, so it becomes important to clearly define audience segments.

“This only takes on more importance for Walmart Connect as it moves into CTV advertising with the Vizio acquisition,” he said. “That’s going to become more prominent quickly here, and so that’s where audience segmentation becomes really valuable.”

The custom targeting can extend beyond categories, Hardy said. For example, brands could aim a health and beauty product designed for health-conscious customers at customers who shopped for natural, gluten-free or antibiotic-free foods.

Next year, a Walmart spokesperson said, the company will use AI to refine the custom targeting and further enhance these recommendations for users, helping them understand what to do next based on their goals.

Lipsman said Walmart’s new AI tools fits a trend within customer insights and data products to integrate AI, specifically querying tools that use natural language.

“The primary benefit of that is just the ability to democratize access to a data tool,” Lipsman said. “Oftentimes within organizations, it’s constrained to a few individual people, and there’s so much value to be derived from data.”