In 2024, Google turned the SERP right into a storefront.
In 2025, it turned it right into a market with an AI-based thoughts of its personal.
Over the previous 12 months, Google has layered AI into practically each inch of the procuring search expertise by merging natural outcomes with product listings, rolling out AI Overviews that change conventional product grids, and introducing a full-screen “AI Mode.”
In the meantime, ChatGPT is inching nearer to turning into a personalised procuring assistant, however for now, probably the most dramatic shifts for SEOs are nonetheless occurring inside Google.
To grasp the affect, I revisited a set of 35,000+ U.S. procuring queries I first analyzed in July 2024.
In immediately’s Memo, I’m breaking down the state of Google Buying SERPs in 2025. A yr later, the panorama seems … completely different:
- AI Overviews have began to displace traditional ecommerce SERP options.
- Picture packs dominate the web page.
- Dialogue boards are on the decline.
Plus, an unique comparability of 2024 vs. 2025 ecommerce SERP options and a full, detailed guidelines of optimizations for the SERP options that matter most immediately (obtainable for premium subscribers. I present you precisely how I do that).
This memo breaks down precisely what’s modified in Google’s procuring SERPs over the previous yr. Let’s goooooo.
Enhance your abilities with Progress Memo’s weekly professional insights. Subscribe at no cost!
Within the final 12 months, Google hasn’t simply reworked itself right into a writer that serves up content material to reply queries proper within the SERP (by way of AI Overviews and AI Mode). It’s additionally constructed out an intensive market for procuring queries.
Nonetheless, Google now offers an entire slew of SERP options and AI options for ecommerce queries which are at the least as impactful as AIOs and AI Mode.
In the meantime, ChatGPT & Co. are beginning to embody product suggestions with hyperlinks, opinions, purchase buttons, and suggestions instantly within the chat. (However this evaluation focuses on Google outcomes solely.)
To raised perceive the important thing traits for Google procuring queries, in July 2024, I analyzed 35,305 key phrases throughout product classes like style, beds, crops, and automotive within the U.S. over the past 5 months utilizing seoClarity.
We’re revisiting that information immediately, inspecting those self same key phrases and classes for July 2025.
The outcomes:
- AI Overviews have began to interchange product grids.
- Ecommerce SERPs are more and more visible.
- There are extra question-related SERP options (like Folks Additionally Ask), much less UGC.
- Fewer movies are showing throughout the SERPs for product-related searches.
Concerning the information:
- This information particularly covers Google search outcomes and options. It doesn’t embody ChatGPT, Perplexity, and so forth. Nonetheless, we’ll contact on this briefly under.
- Over 35,000 search queries had been analyzed, and the identical group was examined in each July 2024 and July 2025.
- The search queries analyzed embody product-related queries throughout a broad spectrum, from model phrases (like Walmart) to particular person merchandise (iPads) and classes (e-bikes).
- In the event you’re curious in regards to the precise checklist of Google procuring SERP options included on this evaluation, they’re included on the backside of this memo.
Earlier than we dig into the findings…
In Google’s shift from search engine to ecommerce market (and from search engine to writer), Google has merged as a lot as potential into the SERP web page.
Net outcomes and the procuring tab for procuring searches had been mixed as a response to Amazon’s long-standing dominance.
The procuring tab nonetheless exists, certain.
However for product-related searches, the principle search web page and the Google procuring expertise look extremely related, with the Buying tab streamlined to a product-grid expertise solely.
In June 2024, I reported in Important SERP Options of Google’s procuring market:
- Google has absolutely transitioned right into a procuring market by including product filters to go looking consequence pages and implementing a direct checkout choice.
- These new options create an ecommerce search expertise inside Google Search and will considerably affect the natural visitors retailers and retailers depend on.
- Google has quietly launched a direct checkout function that enables retailers to hyperlink free listings on to their checkout pages.
- Google’s transfer to a procuring market was possible pushed by the necessity to compete with Amazon’s profitable promoting enterprise.
- Google faces the problem of balancing its function as a search engine with the necessity to generate income by way of its procuring market, particularly contemplating its dependence on companions for logistics.
And now?
Google’s layered AI and customized SERP options into the procuring expertise as properly.
Under are the Google SERP options I’ll be inspecting on this year-over-year (YoY) evaluation, particularly, with a fast synopsis in the event you’re not acquainted.
- Photos: A horizontal carousel of picture outcomes associated to the question pulled from product pages or image-rich content material; normally seem on the high or mid-page and hyperlink to Google Photos or on to supply pages.
- Merchandise: Shows a visible grid or carousel of merchandise with titles, photos, costs, opinions, and retailers. This contains free product listings (natural) and Product Itemizing Advertisements (PLAs) (paid).
- Folks Additionally Ask (PAA): Associated questions customers steadily ask. Clicking a query reveals a supply hyperlink. (These usually inform Google’s understanding of search intent and person curiosity.)
- Issues To Know: An AI-driven function that breaks a subject into subtopics and steadily misunderstood ideas. Discovered totally on broad, academic, or commercial-intent queries, that is Google’s method of guiding customers deeper into a subject and understanding deeper search intent.
- Dialogue and Boards: Highlights related threads from platforms like Reddit, Quora, and area of interest boards. Solutions are sometimes community-generated and genuine. Changed some conventional “Folks Additionally Ask” actual property for procuring or opinions queries.
- Information Graph: Shows structured info about an individual, model, product, or topic-sourced from trusted databases. Seems in a right-hand sidebar or embedded field.
- Shopping for Information: A function that explains what to contemplate when looking for a product, e.g., “What to search for in a DSLR digital camera.” Normally positioned mid-page for commerce-intent queries. It mimics a human assistant or product professional’s recommendation. Comprises snippets and hyperlinks to sources.
- Native Itemizing: Exhibits native enterprise listings with map, scores, hours, and fast name/location hyperlinks. Distinguished in searches with native intent like “shoe retailer close to me” or “espresso outlets in Detroit.”
- AI Overview: Generative AI abstract on the high of the SERP that solutions the question utilizing info synthesized from a number of sources. For procuring queries, it usually contains product summaries.
- Video: A carousel or block of video content material, principally from YouTube, but additionally from different video-hosting platforms. Could embody timestamps, captions, or “key moments” for lengthy movies.
- Reply Field (a.ok.a. Featured Snippet): A direct reply to a question extracted from a single internet web page, proven on the high of the SERP in a stylized field. Usually used for factual or how-to queries. Consists of the supply hyperlink.
- Free Product Listings: Natural product outcomes submitted by way of Google Service provider Middle feeds. These listings present within the Buying tab and sometimes in the principle SERP product grid (distinct from paid Buying adverts).
- From sources throughout the net: A content material block exhibiting opinions or quotes on a product or matter from a wide range of websites. Usually utilized in AI Overviews or product opinions to floor aggregated person sentiment or editorial enter.
- FAQ: An expandable schema-driven block exhibiting widespread questions and solutions sourced from a particular web page. Sometimes seems beneath a website’s natural consequence when FAQ schema is correctly applied.
- PPC: Sponsored hyperlinks proven on the high or backside of the SERP, marked “Sponsored” or “Advert.” These can present up as textual content, product photos/grids, and so forth.
Along with the usual SERP options tracked on this evaluation by way of the above checklist, right here’s a have a look at the present Google procuring market SERP options and/or components (like toggle filters) that we’re coping with on the midway level of 2025.
- AI Mode (Full-Display screen): Interactive, immersive full-page AI procuring expertise with filters and purchase hyperlinks.
- Buying filters inline: Dynamic filters (model, coloration, worth) inside AI Mode and Buying grids.
- Digital try-on: This function was not too long ago launched. It’s a generative AI module exhibiting garments on numerous physique sorts (increasing by class).
- Value monitoring/alerts: Customers can observe worth drops and get alerts by way of Gmail or Chrome. Actually, a reasonably useful gizmo.
- Fashionable shops/high shops: Scrollable carousel of outstanding retailers for the product class.
- Product websites (EU market): Natural function that exhibits outstanding ecommerce domains (on account of regulatory adjustments within the EU).
- Trending merchandise/standard merchandise: Highlights merchandise rising in reputation based mostly on latest search exercise.
- Service provider star scores: Show assessment scores and counts in summaries or tiles.
- Free transport/returns labels: Highlighted callouts in product tiles.
- “Verified by Google” service provider badges: Google-trusted vendor icon in some listings.
- Fast comparability panels: Facet-by-side spec or function comparisons (that is an early-stage rollout, much like Amazon’s product comparability panel or module).
For example with an instance, let’s say you’re in search of kayaks (summertime!).
On desktop (logged-in), Google will now present you product filters on the left sidebar and “Fashionable merchandise” carousels within the center on high of traditional natural outcomes, however beneath adverts, after all.
Picture Credit score: Kevin Indig
Immediately beneath the procuring product grids, you’ve got conventional natural outcomes together with an on-SERP Shopping for Information, much like Folks Additionally Ask questions (which can be included additional down the web page).
Each the Shopping for Information and Folks Additionally Ask options ship solutions with hyperlinks to unique content material.
Picture Credit score: Kevin Indig
On cell, you get product filters on the high, adverts above natural outcomes, and product carousels within the type of Fashionable merchandise or “Merchandise for you.”
Picture Credit score: Kevin Indig
This expertise doesn’t look very completely different from Amazon … which is the entire level.
Picture Credit score: Kevin Indig
Google’s procuring expertise lets customers discover merchandise on a wide range of marketplaces, like Amazon, Walmart, eBay, Etsy, & Co.
From an search engine optimization perspective, the outstanding place of product grid (listings) and filters possible considerably impacts CTR, natural visibility, and in the end, income.
However let’s check out the identical search by way of AI Mode.
Under is the desktop expertise by way of Chrome.
I’ve zoomed out right here so that you get the entire view, nevertheless it takes the person two to a few scrolls to get to the product grid when in an ordinary view.
Picture Credit score: Kevin Indig
Right here on cell, attending to product suggestions takes a number of scrolls. In a single occasion, I acquired a consequence that included an inventory of locations close to me in my metropolis the place I may get a kayak.
Picture Credit score: Kevin Indig
Protecting the present Google procuring SERP expertise in thoughts, right here’s what the information exhibits.
That is probably the most noteworthy shift discovered within the information, as you’ll be able to most likely guess.
Since March 2025, when Google started rolling out AI Overviews extra aggressively, they’ve additionally began changing (natural) product grids.
Picture Credit score: Kevin Indig
The graph above would possibly appear like it represents minimal adjustments while you study it in a timeline view, however you’ll be able to see the development even higher when transferring AIOs to a second y-axis (under).
Picture Credit score: Kevin Indig
I count on AI Overviews to nonetheless present the product grids searchers have turn into accustomed to, though they may take a special kind.
When trying to find [which camera tripod should I buy?], for instance, we discover an AI Overview on the high with particular product suggestions.
Picture Credit score: Kevin Indig
After all, AI Mode takes {that a} step additional with richer product suggestions and shopping for guides.
(Shoutout to The New York Instances and the opposite 5 sources for this AI Mode reply … which now don’t see an advert impression or affiliate click on.)
Picture Credit score: Kevin Indig
On account of this shift, which I predict will solely improve over time, monitoring your model mentions and product hyperlinks in AI Overviews turns into essential. Skip this at your personal threat.
Right here, you’ll see the rise in picture packs over time, with an enormous shift in March 2025.
Picture packs for ecommerce-related queries grew from ~60% in 2024 to a brand new baseline of over 90% of key phrases in 2025.
Picture Credit score: Kevin Indig
Additionally, discover how Google systematically checks SERP layouts between core updates (e.g., the dip within the graph above occurs between the March and June 2025 Core Updates).
Having sturdy product photos, that are correctly optimized, continues to be essential for ecommerce search.
Since January 2025, Google has proven extra Folks Additionally Requested (PAA) options at the price of Discussions & Boards.
Although Reddit is the second most seen website on the internet, I’m shocked to see extra PAA – two years after Google eliminated FAQ wealthy snippets from the SERPs.
Picture Credit score: Kevin Indig
That is one thing you wish to contemplate monitoring for queries which are instantly associated to your merchandise, in the event you’re not doing so already. (You are able to do this in traditional search engine optimization instruments like Semrush or Ahrefs, for instance.)
Since August 2024, Google has systematically decreased the variety of movies within the ecommerce search outcomes.
Picture Credit score: Kevin Indig
Plainly photos have taken numerous the true property movies that used to personal.
Picture Credit score: Kevin Indig
In consequence, movies are much less essential in ecommerce search, whereas photos are more and more extra essential.
In the event you’ve been creating and optimizing movies and haven’t seen the search engine optimization outcomes you wished in your merchandise/website, this could possibly be your sign to spend money on different sorts of content material.
Whereas this evaluation covers Google SERP information particularly, it’d be a miss to not focus on the brand new procuring options in ChatGPT.
Nonetheless, we don’t but have months and months of knowledge on LLM-based conversational product suggestions to present us good, clear info, so I anticipate there might be extra evaluation forward as soon as extra time passes.
ChatGPT’s procuring expertise is beginning to look lots like Google’s – however with a twist: As a substitute of viewing lists of blue hyperlinks or a number of product grids, it curates a conversational shortlist with minimal product listings included.
No affiliate hyperlinks and no paid adverts (but).
Picture Credit score: Kevin Indig
OpenAI integrates real-time product information from instruments like Klarna and Shopify, permitting ChatGPT to floor up-to-date costs, availability, opinions, and product particulars in a shoppable card-style format.
ChatGPT additionally affords a “Why you would possibly like this” and “What persons are saying” generative abstract when a particular product is clicked.
Picture Credit score: Kevin Indig
OpenAI affords the next steerage about how these merchandise are chosen [source]:
A product seems within the visible carousel when ChatGPT perceives it’s related to the person’s intent. ChatGPT assesses intent based mostly on the person’s question and different obtainable context, resembling reminiscences or customized directions….
When figuring out which merchandise to floor, ChatGPT considers:
• Structured metadata from third-party suppliers (e.g., worth, product description) and different third-party content material (e.g., opinions).
• Mannequin responses generated by ChatGPT earlier than it considers any new search outcomes. Be taught extra.
• OpenAI security requirements.
Relying on the person’s wants, a few of these elements might be extra related than others. For instance, if the person specifies a finances of $30, ChatGPT will focus extra on worth, whereas if worth isn’t essential, it could deal with different points as a substitute.
OpenAI additionally explains how retailers are chosen for merchandise [source]:
When a person clicks on a product, we could present an inventory of retailers providing it. This checklist is generated based mostly on service provider and product metadata we obtain from third-party suppliers. Presently, the order by which we show retailers is predominantly decided by these suppliers….
To that finish, we’re exploring methods for retailers to supply us their product feeds instantly, which can assist guarantee extra correct and present listings. In the event you’re fascinated about taking part, full the curiosity kind right here, and we’ll notify you as soon as submissions open.
That being stated, it takes some trial and error to set off product suggestions instantly within the chat.
As an example, the immediate [can you help me find the best kayaks for beginners] ends in an output that features product suggestions, whereas the question [what are the best kayaks for beginners] ends in an inventory with out procuring outcomes, options, or hyperlinks.
Prompts with action-oriented language like “are you able to assist me” and “will you discover” could have a better probability of providing procuring outcomes instantly within the chat, whereas queries like “what’s the finest” and “what are the very best” and “examine the options of” could lead to a wide range of suggestions.
Picture Credit score: Kevin Indig
Featured Picture: Paulo Bobita/Search Engine Journal
