In its latest distribution destination update, Google Cloud notably introduces a radius check function. It also updates the Discovery AI and Recommendation AI services.
In an effort to help the retail industry transform its in-store inventory check processes and improve e-commerce sites, Google Cloud recently announced that it is enhancing its offering for this industry by adding a ray-checking software and updating its Discovery AI and Recommendation AI services. The first is a highly sought-after feature, as lack or lack of inventory is a concern for retailers. According to analysis by NielsenIQ, empty shelves cost major U.S. retailers $82 billion in missed sales in 2021 alone. According to the company, the AI-based verification tool can be used to optimize shelf availability. products on the shelves, provide better visibility of current conditions on the shelves and identify where restocking is needed.
The tool, which is powered by Vertex AI Vision, is powered by two machine learning models – the Product Detector and the Label Organizer – can be used to identify different types of products based on visual imagery and text characteristics, the company said, adding that retailers don’t have to spend time and effort training their own AI models. Additionally, the service can identify products from images taken from different angles and on different devices, such as a ceiling-mounted camera, cell phone or store robot, Google explains in a statement. Images of these terminals are fed into Google Cloud for Retailers. This feature, which is currently in testing and expected to be available globally in the coming months, will not share retailer images and data with Google and can only be used to identify products and labels. added the firm. The French Verteego offers a similar solution, used in particular by Corsica Ferries, Manpower, or even Monoprix to simulate and make reliable its decisions related to stocks, improve the implementation of its promotions at the point of sale, or even manage its prices and the restocking of rays.
Improve the e-commerce experience
To help retailers improve their online product browsing and discovery experience, Google Cloud is also introducing an AI-powered browsing feature in its Discovery AI service. The solution uses machine learning to select the optimal order of products to display on a retailer’s e-commerce site after shoppers choose a category, the firm says, adding that the algorithm learns the order ideal products for each page over time based on historical data.
As it learns, the algorithm can optimize how and what products are displayed for accuracy, relevance, and the likelihood of making a sale. Note that this capability can be used on different pages of a website. “This navigation technology takes a whole new approach, feeding itself, learning from experience and requiring no manual intervention. In addition to dramatically improving revenue per visit, it saves retailers the time and expense of manually managing multiple e-commerce pages,” the company said in a statement. Available to everyone, the function currently supports 72 languages.
Personalized recommendations for customers
To help players create hyper-personalization for their online customers, Google Cloud has released another AI-powered service, within the Recommendation AI offering. The latter is expected to evolve the existing retail search service, Retail Search. It is based on a product pattern recognition machine learning (ML) model that can study a customer’s behavior on a retail website, such as clicks and purchases, to understand the person’s preferences.
The AI then pushes products that match those preferences up the search and browse rankings for a personalized result, the company says. “A shopper’s personalized search and browsing results are based solely on their interactions on that specific retailer’s e-commerce site and are not linked to their Google Account activity,” the Mountain firm said. View, adding that the shopper is identified either by an account they created on the retailer’s site or by a first-party cookie on the website. This feature has been made available to everyone.