AI instruments like ChatGPT and Gemini at the moment are an integral a part of how customers analysis merchandise.
Consumers use them to match choices, consider trade-offs, and make clear selections in the course of the shopping for course of.
However how a lot affect do these instruments really have? And the place do they match alongside conventional serps?
To reply that, we surveyed over 1,000 U.S. customers to learn how they use AI for product analysis. Right here’s what we discovered.
Methodology
To quantify the impression of AI on product discovery and commerce, we surveyed 1,030 U.S. buyers in December 2025 throughout the next demographics:
- Age: Respondents had been comparatively evenly distributed throughout the 18–34 (27.6%), 35–54 (34.7%), and 55+ (37.7%) age brackets.
- Revenue: Respondents spanned all family earnings ranges, from beneath $25,000 yearly (12.8%) to over $200,000 (5.1%). The biggest phase earns between $100,000 and $149,999 (22.2%).
- Training: Half of the respondents (50%) maintain a bachelor’s, grasp’s, or doctoral diploma.
Key Takeaways:
- 48% of U.S. customers use AI each day, and 85% use it not less than weekly. 55% additionally leverage AI particularly for product analysis not less than weekly. AI visibility is not non-compulsory for manufacturers.
- 77% of customers use AI and conventional search collectively. AI has not changed search—it has simply made the journey much less linear.
- 43% of customers have found a brand new model via AI. Nonetheless, rating as the primary model talked about in an AI response barely strikes the needle—solely 20% say a model stands out as a result of it seems earlier within the reply. What issues is how clearly and relevantly a model is described.
- 50% of customers have made a purchase order after utilizing AI throughout analysis, throughout each class and worth level.
- 69% count on AI to play an even bigger position in how they store sooner or later. The shift will not be slowing down.
Let’s dive deeper into these insights.
1. AI Is Now A part of How Customers Store, At Each Stage
AI instruments have moved from novelty to behavior for U.S. customers. 85% of our respondents use AI not less than weekly, 48% interact each day, and 25% use it a number of instances per day
For manufacturers, this implies AI visibility deserves particular consideration as a discovery channel — as a result of their viewers is already there.

Whereas there are lots of AI platforms, shopper AI utilization is concentrated round a couple of dominant platforms. This implies that manufacturers can focus their efforts strategically and don’t have to unfold themselves too skinny:
- ChatGPT leads with 64% of customers utilizing it month-to-month, adopted by Gemini at 49%
- Utilization of Google’s AI Mode (28%) and AI Overviews (22%) means that many customers are utilizing AI inside conventional search quite than switching instruments
- Meta AI reaches 39% month-to-month via Instagram, Fb, and WhatsApp
- Grok (12%), Perplexity (9%), and Claude (8%) path considerably amongst basic customers, although they could index greater amongst skilled audiences

What customers leverage AI for is equally broad. 58% use it to search for info or get explanations, 46% to find out about a subject, 38% particularly for product analysis, and 30% to match choices facet by facet.
For manufacturers, this implies there are alternatives to seem in AI responses throughout the whole purchaser journey — from after they begin to discover a subject to the second they resolve between manufacturers.

2. Customers Depend on AI to Analysis, Evaluate, and Consider Merchandise
That final level exhibits AI is not only getting used on the prime of the funnel—U.S. customers are using it throughout each stage of the acquisition resolution:
- 57% use AI to slender down their selections
- 51% use it throughout early discovery
- 53% examine merchandise they’re already contemplating
- 50% use it to make a last resolution, with 55% doing this not less than weekly, and 25% each day.
As well as, 55% use AI to analysis merchandise not less than weekly—25% of whom accomplish that each day.

When utilizing AI in these methods, the queries customers run are extra focused than a typical search:
- 52% specify constraints upfront—a finances, a required characteristic, a compatibility want, or a selected use case
- 33% commute with the AI, refining their query a number of instances
- Solely 43% begin with a broad question earlier than narrowing down
This stage of specificity implies that a transparent and exact product description can instantly have an effect on whether or not AI surfaces it in its solutions to customers.

With regards to making a purchase order, nonetheless, AI acts as a filter quite than a better:
- 47% of customers use AI to slender down their choices, however then flip to Google for opinions or pricing
- 40% use AI to analysis or examine choices, however full the acquisition elsewhere
- 39% ask AI for suggestions, after which store on marketplaces like Amazon
- Solely 20% say they use AI all through the whole course of, from analysis to last resolution

What this means is that AI determines which manufacturers make the shortlist, however different channels nonetheless matter for buy selections.
3. AI and Google Search Are Not Competing—Customers Use Each
Regardless of what many assume, AI has not changed conventional search. Somewhat, it enhances it:
- 77% of our respondents use AI and serps collectively, and solely 4% rely totally on AI
- Search engines like google and yahoo like Google stay the extra frequent place to begin at 33%, with AI shut behind at 26%
- 18% swap between each all through the analysis course of
In different phrases, customers are utilizing each channels for a similar buy resolution.

This implies manufacturers want content material that each ranks effectively on Google and surfaces in AI responses.
What’s extra, AI solutions aren’t the tip level for customers’ exploration: 94% of customers click on hyperlinks in AI responses not less than generally, with 38% saying they accomplish that usually or nearly all the time. Solely 3% say they by no means click on.
For manufacturers, this means that AI citations is usually a highly effective supply of high-intent visitors.

The truth is, we discovered that AI summaries usually speed up shopper analysis with out changing web site visits:
- 87% of U.S. customers say AI summaries assist them perceive manufacturers quicker
- 68% go to model web sites simply as usually or greater than earlier than, with 22% extra more likely to go to model websites after an AI interplay
- Solely 11% say they’re much much less more likely to go to a web site after utilizing AI
In different phrases, AI is compressing the early levels of analysis, not eliminating what comes after.

4. AI Is Altering How Customers Uncover and Consider Manufacturers
A point out in an AI response is not only passive publicity—it drives speedy motion for customers.
When an AI device mentions a model or product, 40% of customers search Google for extra info, 36% use Google to match it with options, 34% ask the AI follow-up questions, and 28% go on to the model’s web site.
Solely 8% ignore the point out except they already know the model. For manufacturers not at the moment showing in AI responses, that is the pipeline they’re lacking.

In different phrases, being found via AI is already a profit for manufacturers: 43% of customers report having discovered a brand new model via an AI device, whereas 47% of U.S. customers report noticing AI-mentioned manufacturers usually or fairly often.

For manufacturers that do seem in AI responses, the place of the point out issues lower than the precise content material. Solely 20% say a model stands out as a result of it seems greater or earlier within the reply.
What really influences their consideration is readability concerning the model:
- 43% say a clearer or extra detailed rationalization makes a model stand out
- 39% be aware of the value or worth context
- 37% are influenced by descriptions which are a superb match for his or her particular wants
- 28% are influenced by direct comparisons between choices

The implication for manufacturers is {that a} particular and high-quality description in AI responses—incorporating their very own content material, opinions, and third-party protection—issues greater than the place they’re talked about within the response.
5. Customers Belief AI Suggestions, However Confirm Earlier than Shopping for
Customers belief AI sufficient to start out their analysis with it, however not sufficient to skip verification earlier than making a purchase order resolution.
75% of U.S. customers fee their belief in AI suggestions at 3 or 4 out of 5. Solely 20% belief AI fully, and solely 7% categorical low belief.
.png)
That conditional belief interprets into constant verification conduct, the place customers will double-check AI’s suggestions elsewhere earlier than shopping for.
The truth is, 86% of U.S. customers say they confirm AI model suggestions not less than generally, together with 20% who all the time accomplish that.
Google is the first validation channel at 68%, adopted by model web sites at 48%. Evaluation websites (35%), YouTube (35%), pals or household (33%), and social media (30%) additionally play vital roles.

For manufacturers, because of this being talked about in AI is simply the beginning of the client’s journey. Your presence on Google, your web site, and third-party evaluation platforms all want to carry up as soon as a shopper goes wanting.
6. AI Already Influences Actual Purchases Throughout Each Class
AI is already producing conversions for manufacturers, and never simply low-stakes purchases. It’s driving actual purchases throughout each class and worth level.
50% of customers have made a purchase order after utilizing AI throughout analysis, whereas 22% have accomplished a purchase order instantly inside an AI device.
And whereas that native AI commerce will not be but dominant, it’s actual and rising.

Customers are utilizing AI for on a regular basis and high-consideration purchases alike:
- 39% say AI has influenced a retail or shopper items buy
- Leisure and life-style purchases, comparable to meals (29%), wellness (29%), and electronics (27%) are among the many most typical classes
- Larger-consideration services and products comparable to journey (21%), training (16%), dwelling companies (15%), and monetary companies (13%) are much less frequent, however nonetheless present vital AI affect

By way of the worth and impression of their purchases, 37% of U.S. customers additionally depend on AI most for mid-range purchases like electronics and subscriptions, 28% use it for high-cost or high-risk selections, and 36% say they use AI equally throughout all buy varieties.
No class is insulated from AI affect at this level.

And right here’s what issues essentially the most: this behavioral shift will not be slowing down.
69% of U.S. customers count on AI to play an even bigger or a lot larger position in how they store sooner or later. Solely 3% count on it to shrink.
On the similar time, 46% anticipate relying much less on conventional serps as AI improves, and 42% much less on social media. Evaluation websites (34%), influencers (33%), advertisements (30%), and blogs (28%) are additionally anticipated to lose floor.

For manufacturers nonetheless treating AI visibility as non-compulsory, that window is closing. It’s turning into an crucial.
Future-Proof Your Model for Fashionable Search Discovery
Customers are utilizing AI proper all through the entire shopping for course of: to be taught, to match and slender their choices, and to validate their selections.
Whereas they nonetheless flip to serps and model web sites, AI is more and more shaping which manufacturers they contemplate within the first place.
What this implies for you and your model is that each AI visibility and conventional search visibility are essential for succeeding.
Semrush One may also help you observe and optimize your model’s efficiency throughout all search surfaces. It exhibits how your enterprise seems in AI responses, helps you observe competitor strikes, and divulges how Google and LLMs place your model.









