Google Shopping receives over 1.2 billion searches every month, with the number of clicks growing by 17.7% annually. It is also where 36% of product discovery happens.
This shows how important it is as a channel for revenue and brand awareness.
As the new Google Shopping gets rolled out across the world, the shopping experience gets increasingly attractive and intuitive to users.
And while there’re competing channels like Perplexity Shop like a Pro or social commerce, with ecommerce sales projected to grow at an average of 7.6%, the number of searches on Google Shopping is likely to increase over time.
How does the new Google Shopping work?
Similar to the old Google Shopping, it features products pulled from marketplaces and ecommerce stores. Each product comes with an image, product title, pricing, reviews, tags, etc.
Ads still show up prominently at the top. Filters appear on the left of the page, allowing you to narrow down your searches.
But there are a few differences.
1. AI intelligently recommends the most relevant products
When you search for products on the new Google Shopping, you’ll now see an AI summary sharing what to consider, along with articles that explain more in detail.
In few occasions, the AI summary features top 3 recommended items like this.
The AI-generated information don’t always appear. We found that they mostly only show up when we search for something less direct. For example, “puffer jacket for women” doesn’t deliver AI summaries, while “puffer jacket for winter” or “puffer jacket for winter in Korea” do.
The former is more straightforward. While the latter needs some uncovering (i.e. which puffer jackets are warm enough for the cold temperatures in Korea?).
Below this AI summary lies a much more organized view of the long list of products. The products are categorized with category titles and descriptions that explain why they work for you based on your query.
Here, you can see the category title and description matching what I searched for, emphasizing the harsh winter of Korea.
2. Get more personalized results and feed
Before you start searching, you’ll see your feed showcasing a list of products. These products are recommended based on your preferences and shopping habits.
The same goes for the results you get from searches—they’re also personalized for you.
But you can set your own preferences or turn it off altogether if you prefer more control or have a neutral feed.
3. Find the best prices with Google’s deal-finding tools
On the top right of Google Shopping, you’ll notice the word “Deals”. Click on it and you’ll be brought to a personalized deals page where you can browse deals tailored to you. These deals are aggregated from ecommerce stores everywhere.
How to optimize for the new Google Shopping?
Understanding the Google Shopping Graph
Optimizing for Google Shopping starts with understanding how the Google Shopping Graph works.
The Google Shopping Graph is an AI-driven, real-time dataset that powers Google Shopping’s search and discovery features. It processes billions of product listings globally, using machine learning to match shoppers with the most relevant products.
Google pulls product data from multiple sources, including:
- Google Merchant Center feeds (data provided by merchants).
- Retailers' websites (via structured markup or crawling).
- User-generated content (reviews, ratings, and even videos).
Let’s say someone searches for something very specific, like this example Google gave:
Maybe you’d love a women’s red puffer coat that’s cropped, shiny, and has a fleece hood. Size medium. And you want it on sale, with free shipping. That criteria is getting complex. What if you also want the coat to be suitable for extreme weather and lightweight?
The Shopping Graph processes billions of listings in real time, combining product information with data from the web—such as images, descriptions, reviews, and even YouTube videos—to refine search results.
In the red puffer jacket example, the Shopping Graph interprets search intent and cross-references details like:
- How well the jacket performs in cold temperatures (insulation level, wind resistance).
- Whether it’s lightweight and packable (material, thickness).
- Sustainability aspects (eco-friendly fabrics, recycled materials).
By understanding these nuanced characteristics, Google ranks and recommends products that best fit the shopper’s specific criteria—making search results more personalized and relevant.
Since Google Shopping Graph takes information about listings from GMC and across the web, ecommerce brands should work on optimizing listings on GMC in a few ways.
Optimize your listings on Google Merchant Center
On GMC, there are mandatory attributes like:
- Product ID
- Product title
- Product description
- Product availability
- Product’s landing page link
- Product image link
- Price
- Identifiers like GTIN / MPN
- Brand
The first thing that people see are your product images, so ensure that your images are of high-resolution and backgrounds help accentuate the product features (our AI bulk image editor is great for that). Titles should also be clear and include keywords that your target audience search for. Good reviews are essentials.
Then there are also optional attributes like product highlight, product details, color, size, gender, and more here.
With the new Google Shopping, it gets increasingly important to include as many product attributes as possible, even the optional ones.
This helps Google understand your product better and recommend your products, especially for more nuanced and specific searches like “jackets for December in Tokyo”.
Here, we see that Google begins by researching how the temperature is like in Tokyo in December, and whether it’s windy or rains, before deciding on the type of products to recommend and sharing options that match the information it gathers.
This means that attributes, descriptions, product details are important. They need to be highly detailed, accurate, and contain keywords that searchers use. Which means understanding your target audience is crucial.
And because the research starts from assessing the weather and ideal products to use, the information you put out there matters. Which brings me to my next point.
Tip: Use Hypotenuse’s AI product attribution feature to enrich your products with accurate, high-quality and up-to-date attributes in bulk.
Publish high-quality content to shape the narrative
While you optimize your product page content and the attributes you share with GMC, it’s important to also create other forms of content like product guides and blog articles.
REI does this very well. Along with keeping their product pages detail-rich, they also create guides on how to choose products, what to bring on trips, and more.
This gives them a much better chance at ranking on Google Shopping. And you’ll see why.
In this example, when we searched for “women’s hydration bags good for day hikes”, REI’s article on how to choose hydration packs appeared in the AI summary. And along with that, their product shows up right below the summary.
REI showcased thought leadership and guidance on how anyone should choose a hydration bag if they’re going on day hikes, which later led into the product they recommended. Especially if there are few resources online, such guides can easily capture a spot in the AI summary.
This also comes with understanding your target audience inside out and what information they need to make a purchase.
Understand your target audience and how they search
There are many factors that shoppers consider before buying something. A shopper looking for a dining table might think about how it’d look in their new home. Someone starting their fitness journey might not know how a pair of tights should fit. Another shopper new to the corporate world might need help getting their business casual outfit sorted.
These are questions you should answer either in your product page or other content you publish.
A few way you can research and understand your target audience better:
- Hear from your customers through Customer Service queries
- Review competitor and marketplace listings
- Check out People Also Ask and Google autocomplete suggestions
- Look at communities like Reddit and what they discuss
- Browse social media like Pinterest and Instagram to understand what is being recommended
For terms that appear consistently, use them in your product content such as titles, descriptions, or FAQs. Create product guides to support users throughout their shopping journey—from research to purchase and post-purchase.
Besides that, you can also refer to popular guides put out by non-ecommerce brands. Then, tailor your product pages to match these.
- What do they recommend?
- What terms do they use?
- What are specific needs they address?
Some of these guides and review blogs are also good opportunities to showcase your brand.
Monitor and work on brand mentions
Brand mentions are critical for brands to rank in AI and organic search.
With the new Google Shopping featuring AI summaries, it matters even more.
There are a few types of pages that Google reference:
- Ecommerce blog articles
- Industry guides
- Affiliate articles
- Community posts and discussions
Partner up with industry blogs on listicles or engage affiliates for product reviews.
Be sure to research the following before reaching out to ensure your products are well represented:
- Audience fit: Ensure the blog or affiliate site’s audience aligns with your target customers. Look at demographics, interests, and past engagement metrics.
- Previous content and review style: Analyze past listicles or product reviews to see if their tone, depth, and accuracy align with your brand messaging.
- Brand competitor mentions: Check if they’ve featured direct competitors. If so, identify ways to differentiate your product or secure better positioning.
- Engagement & Audience Trust: Look at the comments, social shares, and reader feedback on past product reviews to gauge credibility and influence.
- Domain authority & SEO strength: Use tools like Ahrefs, Moz, or SEMrush to check their domain authority (DA) and organic traffic. Prioritize sites with strong rankings and relevant keywords.
- Backlink quality & referral potential: Ensure the site offers do-follow links and drives high-intent traffic rather than low-quality, spammy referrals.
Additionally, engage with relevant communities like Subreddits where your audience spends time.
Follow trends and adjust your content
In some industries like fashion and apparel, trends come and go very quickly. Besides launching new collections, your content should evolve together with it.
Stay on top of trends by:
- Reviewing Google Trends for news related to your industry
- Discovering new keywords you’re ranking for on Google Search Console
- Following industry blogs, social media accounts, and relevant online communities
Pick out new keywords that you can use on product pages, create new categories under trending terms, and update attributes on GMC as well.
The new Google Shopping filters also give you clues on what’s popular as well.
For example, under the search term “women’s brunch outfit”, you can see popular patterns like leopard, striped, or plaid, and silhouettes like wrap, fit and flare, and asymmetrical. If these are relevant, add them to your product attributes.
Conclusion
Google Shopping is always changing, and keeping up with the latest updates is key to getting your products in front of the right shoppers.
With the New Google Shopping, there are even more ways to get discovered—thanks to AI-driven recommendations, better product search, and a more personalized shopping experience. That means how you optimize your listings matters more than ever. High-quality product data, engaging descriptions, and well-structured attributes can make all the difference in getting clicks and conversions.
That’s where Hypotenuse AI comes in. Instead of spending hours tweaking product descriptions or structuring data manually, you can automate and optimize everything at scale—from content generation to enrichment. With AI-powered insights, you’ll always be ahead of the game, ensuring your listings stay competitive and perform at their best.
