Blog post

Walmart’s Intelligent Retail Lab Tells No Tales

By Robert Hetu | October 07, 2022 | 0 Comments

AIRetail Digital Transformation and Innovation

A recent visit to the Walmart IRL (Intelligent Retail Lab), located in a neighborhood market store in Levittown, NY, provided some useful insights and prompted some questions. The store was converted to an artificial intelligence laboratory with a great deal of fanfare in 2019. Now, almost three and a half years later, I wondered about the results from the experiment.  With my fellow Gartner analyst Sandeep Unni, we took a spur of the moment chance to visit. I searched online for any recent articles on the Intelligent Retail Lab but there really hasn’t been any coverage since early 2020.  However, while researching I did learn that, in popular vernacular, IRL means in real life, so there is that. I did play around with the Point Cloud interactive display at the Intelligent Retail Lab. 

Here I Am IRL, Or At Least IPC

Walmart's Intelligent Retail Lab has and interactive Point Cloud technology display
Robert Hetu IRL

What Stands Out From Intelligent Retail Lab Today

  • There are well over a thousand cameras installed overhead, but customers seem to take this in stride.
  • Its very well staffed with busy associates in nearly every aisle.
  • The shelves are all front faced and orderly.
  • There are many holes in the planograms where stock outs were obvious, but given the condition of the store and the staffing available, these could not be due to in store operational issues.
  • The checkouts are completely uncluttered, save for some low profile drink coolers.  There are none of the typical impulse products like candy or gum anywhere in the checkout area.
  • Trying to compare the store from three years ago to now, the only significant changes seem to have been applied in the checkout area, which heavily focuses on self checkout, and prominently features a single checkout lane that appears to have more experimental equipment.
  • There is an interesting interactive display, although not prominent, that illustrates how computers use Point Cloud technology to understand 3d depth and keep products in stock. 

My Questions

  • Why exactly is there no use of impulse products at the checkout stands at IRL? My companion wanted to purchase a pack of gum and we had to go back into the aisles to look for any.  I wonder if the choice was purely aesthetic, or if it had something to do with the computer vision capabilities. 
  • What is the purpose of the new checkout stand? It appears to scan many items all at once, perhaps from a small shopping basket. It was not operational when we visited but an associate did say that it has been in use.
  • Why are there so many out of stock items in the planograms? There was obviously a lot of staff and the maintenance of the shelves for what was in stock was nearly perfect. I suspect there are ongoing supply issues but it did make me wonder if there was not something else that could have been directed into those locations at least temporarily.
  • Why hasn’t there been more obvious changes? Now, there may have been a lot of assortment edits and planogram changes, but overall it seemed pretty static.  
  • What is the story that the intelligent retail lab can tell? I am a little surprised that its been low key for so long. Perhaps intentionally by Walmart, the lack of publications and press may reflect a good amount of deep learning that has been repurposed into the Walmart strategy. It could also mean that there is not much of a tale to tell. 

The Gartner retail team is fully immersed in research around AI in retail, and the role of the store as the hub of digital and physical commerce.  Here are a few of my recent blogs on the topic:

Stores Are Not Out Of Fashion

Retail 2022: Betting Big On Digitalized Stores

The Gartner Blog Network provides an opportunity for Gartner analysts to test ideas and move research forward. Because the content posted by Gartner analysts on this site does not undergo our standard editorial review, all comments or opinions expressed hereunder are those of the individual contributors and do not represent the views of Gartner, Inc. or its management.

Leave a Comment