AI for Retailers: Google Launched New Features Based on AI for Retail Websites

TL;DR

Google Cloud launched Discovery AI for retail, bringing machine learning directly to e-commerce websites to personalize shopping experiences and automate inventory management through visual recognition technology.

  • Why it matters: Retailers can reduce search abandonment and boost conversions with AI-powered personalization that understands user behavior and browsing history.
  • How it works: The system uses image attributes and previous browsing data to create dynamic product recommendations tailored to individual shoppers.
  • The big picture: AI shopping assistants now offer real-time suggestions, mimicking trusted salespeople while learning from customer interactions.
  • Case in point: Macy's already uses these solutions to deliver individualized online shopping experiences for customers.
  • What's next: Google's shelf check feature will use machine learning to identify billions of products and verify in-store inventory through camera systems.

Recently, SEO professionals have been trying to understand the impacts of the mass arrival of AI tools, such as many AI-writer assistants (e.g. Lex and ChatGPT).

They need to rethink strategies, considering all these AI tools‘ advantages but making sure to maintain humanized experiences.

And if that wasn’t enough to worry about, well… there’s more to come: Google Cloud has just launched new features also based on AI and machine learning as enablers of retail websites. 

Google calls this bundle of features “Discovery AI for retail”, intending to enable e-commerce sites to offer a more personalized experience in search results and Google recommendations inside their properties.

These solutions have the primary purposes of reducing search abandonment, improving the shopping experience, and increasing conversion.

Note that I mentioned “inside its properties” above. That’s right! I’m not referring to the SERP results, but literally to the browsing experience that e-commerce sites themselves may offer.

How’s that going to work? Well, allow me to explain everything you need to know about AI for retailers. 

Neste post 2

Improved Shopping Experience

Google Cloud, which announced the launch of these features on January 13, highlights that this set of solutions provides a significant improvement in the shopping experience using features such as:

  • Image attributes that will help find products
  • Understanding user behavior through previous browsing experiences.

A highly likely potential scenario is that retailers may offer consumers more personalized product lists that closely identify with their real expectations.

What we usually see nowadays are lists of products classified in ways that are much more focused on business needs, with highlights such as “bestsellers” or collections of products according to seasonality, season changes, etc., which do not always represent what the user wants to buy.

Product recommendations will also have a more dynamic presentation, which can be much more efficient and, consequently, increase the chances of sales.

Some retailers are also experimenting with AI-generated shopping assistants—little chat widgets that do more than just answer questions. They offer actual suggestions in real time, attempting to mimic the trusted, friendly salesperson you rarely find in physical stores anymore. I’m not saying these bots are perfect (they get stumped every now and then), but when they work, it’s hard not to be at least a little impressed.

On a related note, there’s a bit of a double-edged sword situation happening as AI gets cozier with retail. The more sites know about your browsing and buying habits, the creepier things can feel if you’re not braced for it. But if you can get past that initial “how did they know I wanted those sneakers?” moment, the tailored experience tends to keep folks coming back—and shopping more often, at least if you believe the numbers coming out of 2025 so far.

See Macy’s case, which already uses the solution, making use of AI for retailers and offering more individualized experiences for its customers in the online store.

Macy’s delivers a personalized shopping experience with Google Cloud Discovery solutions (Source: Google Cloud YouTube channel)

In addition, I also believe we can expect better results for “long tail” searches because AI can better understand them. Until then when we used to look for a specific product, these sites generated poor returns.

AI-based Inventory Control

Another huge feature that Google Cloud includes in their AI offerings, is “shelf check”.

Last Friday, Google CEO Sundar Pichai posted on his Twitter account that one of these new AI features would also “use machine learning models to identify billions of products based on visual and text features, helping retailers check their in-store shelf stock”.

On the same day, in an interview with the Wall Street Journal, Google Cloud said that this feature should be available in the coming months.

Reinforcing Pichai’s post, they also claim that the algorithm will be able to detect and verify the availability of packaged products still on the shelves through the images captured by the cameras installed in the physical storage locations.

The solution would solve a current problem for retailers who currently control availability updates manually with little security, and who do not always carry them out in real time.

Frequently Asked Questions

How does Google's Discovery AI improve product recommendations for shoppers?

Discovery AI analyzes user behavior patterns and browsing history to create personalized product lists that match individual preferences. Unlike traditional bestseller lists or seasonal collections, these recommendations adapt dynamically based on what each customer actually wants to buy. The system processes image attributes and previous interactions to suggest products that align with real shopping intent. Test this technology by exploring e-commerce sites that have implemented Google's retail AI solutions.

What makes AI-powered shelf check different from manual inventory management?

Google's shelf check uses machine learning models to identify billions of products through visual and text features captured by in-store cameras. This automated system provides real-time inventory verification, eliminating the security gaps and delays of manual stock updates. Retailers can detect product availability instantly without human intervention, ensuring accurate shelf monitoring across multiple locations. Evaluate your current inventory processes to identify where visual AI could reduce manual oversight.

Can AI shopping assistants actually replace human sales expertise?

AI shopping assistants offer real-time product suggestions and answer customer questions, but they work best as supplements to human expertise. While these chat widgets can get stumped occasionally, they excel at processing vast product catalogs and customer data to provide instant recommendations. The technology mimics friendly salespeople by learning from interactions and adapting suggestions based on individual preferences. Experiment with AI assistant features on retail websites to understand their current capabilities and limitations.

How do retailers balance personalization with customer privacy concerns?

The challenge lies in leveraging browsing and buying habits for better recommendations while maintaining customer trust. Many shoppers initially feel unsettled by highly targeted suggestions but tend to appreciate the tailored experience over time. Retailers must be transparent about data usage and provide clear opt-out options for customers who prefer less personalized experiences. Review your favorite retail sites' privacy policies to understand how they handle personalization data.

MM Matt Montenegro