AI chatbots person opened up a full caller customer acquisition funnel. And marketers now person to grapple pinch the nonstop aforesaid questions of assets allocation that dictate hunt motor budgets: prioritizing optimization for integrated discovery, aliases paid advertising.
Exploding Topics wanted to find retired really consumers are really researching products online, to thief marketers and marque strategists make much informed decisions astir their AI optimization strategy. So successful our biggest study to date, we asked complete 2,300 group successful the U.S. really they usage AI successful their purchasing journeys.
The image that emerged was 1 of an AI gatekeeper softly swinging user acquisition decisions.
Chatbots person go a large portion of the mean consumer's merchandise research, and customers are discovering caller brands done AI. Consumers are researching section and large-scale businesses alike, and their AI usage is expanding complete time. Almost two-thirds of the AI users we surveyed person astatine slightest partially replaced product-related Google searches pinch AI conversations.
But arsenic good arsenic being an avenue for marque discovery, AI chatbots are besides talking would-be customers retired of purchases. More than half of AI users person decided not to acquisition thing connected the ground of accusation from a chatbot.
At the aforesaid time, much than 4 successful 10 group actively dislike AI chatbot ads, and the mean consumer’s sentiment of a marque is somewhat much apt to worsen than amended aft seeing an advertisement for it. So chatbot advertisement abstraction is not needfully the easy hole to AI visibility issues.
Instead, our results show the existent and coming value of Generative Engine Optimization (GEO) efforts. Keeping connected way of your AI marque beingness isn’t conscionable a luxury for securing other customers, it’s an progressively basal prerequisite for ensuring your income aren’t getting intercepted.
Fast facts
Among each consumers:
- Roughly 1 successful 5 would prime an AI chatbot if they could only usage 1 merchandise investigation tool
- Almost half (47.54%) sometimes consult an AI chatbot to get seller accusation earlier making a purchase
- 74.15% would beryllium astatine slightest somewhat little apt to bargain a merchandise if an AI chatbot flagged mixed aliases antagonistic reviews
Among each AI users:
- 65.01% have astatine slightest partially replaced Google searches pinch AI chatbots for merchandise research
- 71.63% feel neutral aliases antagonistic toward ads successful AI chatbots
Among those who usage AI astatine slightest weekly:
- 73.58% have bought a merchandise based connected an integrated AI recommendation
- 81.24% believe AI chatbots are tin of recommending the champion brands and products
- Almost 7 successful 10 (69.7%) person decided not to bargain thing connected the ground of AI advice
AI chatbots are the caller influencers
Organic AI mentions person an progressively decisive effect connected user activity.
41% of consumers person made a acquisition based explicitly connected an AI chatbot recommendation. Among each consumers who usage AI astatine slightest occasionally, that fig rises to 57.51%.
And the image is moreover much stark among AI powerfulness users. 80.45% of consumers who usage AI aggregate times per week person bought an AI-recommended merchandise astatine slightest once, pinch much than half (54.95%) doing truthful aggregate times.
Crucially, the persuasive powerfulness of AI chatbots useful successful some directions. 57.5% of AI users person decided not to bargain thing based connected accusation provided by an AI chatbot.
When you select further to only see the consumers who named AI chatbots arsenic 1 of the sources they usage erstwhile making a acquisition online, 80.93% of group person decided against a acquisition connected the ground of AI guidance. 42.81% person done truthful aggregate times.
Looking astatine all consumers again, 74.15% would beryllium astatine slightest somewhat little apt to bargain a merchandise if an AI chatbot flagged mixed aliases antagonistic reviews. Among existent AI users, 85.31% would rethink a acquisition if a chatbot revealed unfavorable reviews.
Knowing that AI chatbots tin dissuade users from purchases almost arsenic overmuch arsenic they tin seduce them underlines the reality that GEO is nary longer optional. Marketers who aren’t optimizing for AI aren’t conscionable missing retired connected imaginable acquisitions, they whitethorn beryllium having would-be purchases actively intercepted by chatbots.
AI mentions person tangible marque impacts
AI recommendations tin do much than power a azygous purchase. They tin present brands and products to full caller audiences.
59.27% of each AI users person discovered a marque aliases merchandise they were antecedently unaware of done an AI chatbot recommendation. More than a 4th (25.79%) person “frequently” discovered caller brands and products.
(These findings tally with Semrush's purchaser travel study from March, which recovered 43% of consumers person discovered a caller marque done an AI tool. Across each consumers, our study puts the aforesaid fig astatine 43.14%.)
Among consumers who are actively utilizing AI for merchandise research, only 9.38% person ne'er discovered a caller merchandise aliases marque from a chatbot recommendation.
Brands guidelines to use massively erstwhile AI chatbots present them to caller customers. More than half of each consumers judge AI chatbots are tin of recommending the champion products and brands.
And erstwhile you attraction connected the consumers who presently usage AI, 70.62% judge that chatbots tin urge the champion brands and products. That rises supra 80% for those utilizing AI astatine slightest weekly.
In different words, a important mostly of the immense and increasing AI chatbot personification guidelines believes that the exertion tin supply valuable and actionable brand/product recommendations. Looked astatine done this lens, optimizing for AI marque mentions seems progressively non-negotiable for marketers.
As good arsenic discovering caller brands done integrated mentions, consumers are besides utilizing AI chatbots to actively investigation brands they are already alert of.
Almost half (47.54%) of each consumers now astatine slightest occasionally consult AI chatbots to get accusation astir a institution aliases seller earlier making a acquisition decision. Brands now person to relationship for the reality that location is simply a adjacent 50-50 chance that a would-be customer mightiness see moving an AI “background check.”
Excluding AI non-users, 65.42% of consumers study sometimes consulting an AI for institution accusation earlier committing to a purchase. Among these AI users, 32.68% “frequently” consult AI chatbots for accusation astir a seller.
Sure enough, astir of the consumers who person ever researched a institution utilizing an AI chatbot person done truthful recently. 48.97% person researched a institution utilizing AI wrong the past week, and a further 20.3% person done truthful successful the past 2 weeks.
So what will AI opportunity astir your marque erstwhile your customers ask? Coupled pinch our findings connected the influential domiciled of chatbots arsenic imaginable acquisition interceptors, this looks to beryllium an progressively basal mobility for marketers to answer.
The powerfulness of AI chatbots for section businesses
Local, beingness businesses mightiness deliberation that AI is of constricted relevance to them. As it turns out, customers deliberation very differently.
In fact, 87.41% of group who usage AI for company/seller investigation would astatine slightest occasionally consult an AI earlier hiring a section business.
Almost a 3rd (31.48%) of consumers who usage AI for institution investigation would ever consult AI chatbots earlier hiring a section business.
In galore cases, consumers judge AI chatbots are amended astatine giving proposal connected section businesses.
42.01% of group who usage AI to investigation companies/sellers deliberation that AI gives proposal connected section businesses amended than it recommends large-scale companies. A further 43.7% judge AI handles section business proposal conscionable arsenic good arsenic it handles large brands.
For business owners, the instruction is clear: sloppy of your standard and scope, galore consumers will usage AI to investigation your company. Taking power of the responses they person tin person a tangible effect connected your bottommost line.
Consumers move to AI en masse for merchandise research
Discussion astir user take of AI is nary longer a speculative topic: chatbots person already been adopted astatine standard crossed astir demographics.
33.23% of group usage AI chatbots for illustration ChatGPT, Gemini, Claude, and Perplexity aggregate times per week. A further 15.41% usage chatbots weekly.
There is still a important number (29.96%) who do not usage AI chatbots astatine all. But AI non-usage is chiefly confined to the over-60s, among whom 60.05% ne'er usage AI chatbots. Only 13.61% of each consumers aged 18-60 study ne'er utilizing AI chatbots.
Likewise, 65.5% of consumers aged 18-60 usage AI chatbots astatine slightest erstwhile per week.
And erstwhile a user is simply a regular AI user, it’s apt that they usage AI for merchandise research.
55.13% of respondents who usage AI astatine slightest erstwhile per week besides said they usage AI chatbots arsenic a root of accusation erstwhile making a acquisition online. Among these regular users, AI is simply a much celebrated merchandise investigation instrumentality than friends and family (49.28%), societal media (48.38%), unit websites (44.42%), and YouTube (43.35%), though hunt engines and reviews still lead.
Across each demographics, including AI non-users, a much humble 29.09% of consumers study utilizing AI chatbots for merchandise research. But much than 60% of respondents expect they will beryllium utilizing AI for merchandise investigation much often successful 2 years’ clip if the exertion continues to beforehand astatine a akin pace. So marketers should beryllium treating GEO arsenic a imperishable and progressively important strategical pillar.
AI is besides steadily increasing successful prominence arsenic a first larboard of telephone for merchandise research.
Excluding complete AI non-users, 12.72% of consumers now commencement their online purchasing travel pinch AI chatbots. Search engines (29.89%) and online reviews (28.13%) are still much communal starting points, but AI is the third-most-common spot for consumers to statesman their research.
AI users pinch upper-middle family incomes are much apt connected mean to commencement their purchasing journeys pinch an AI chatbot. 25.12% of AI users pinch a family income of $125,000-$199,999 move to AI chatbots first, narrowly beating online reviews (24.13%) and only trailing hunt engines (35.58%).
Interestingly, the percent of AI users who commencement online purchases via chatbots drops backmost down to 10.19% among the very highest earners ($200,000+ family income). That’s comparable to little earners; only 8.8% of users pinch family incomes from $0-$124,999 statesman purchasing journeys by consulting chatbots.
Nonetheless, consumers aren’t conscionable starting their merchandise investigation pinch AI. Almost 1 successful 5 group (19.43%) would take an AI chatbot if they could only prime 1 source earlier making a purchase.
Filtering retired those who don’t usage AI chatbots astatine all, much than a 4th of consumers (27.37%) would opt for chatbots arsenic their sole root of merchandise investigation if made to choose. Meanwhile, 44.39% of AI “power users” (multiple uses per week) would prime AI chatbots arsenic their only merchandise investigation tool, versus 26.29% for hunt engines and 22.59% for reviews.
AI chatbots situation Google supremacy
Consumers are adopting AI chatbots for merchandise investigation pinch specified enthusiasm that devices for illustration ChatGPT are gradually opening to eat into Google’s near-monopoly.
More than 1 successful 5 AI users person wholly replaced their product-related Google searches pinch AI chatbots. And 65.01% of consumers who usage AI person astatine slightest partially swapped retired Google for chatbots.
AI users aged 45-60 are astir apt to person phased retired Google wholly for merchandise research, pinch 39.04% reporting that AI chatbots person “completely” replaced the hunt engine. By contrast, 63.36% of AI users aged complete 60 study that they person not replaced immoderate product-related Google searches pinch AI chatbot conversations.
Despite galore users swapping immoderate Google searches for AI queries, the mean user has not reduced their wide usage of Google for acquisition investigation successful the past year. Almost half (47.6%) of AI users person astir maintained their level of Google usage successful the past 12 months, while 23.01% study utilizing the hunt motor overmuch more.
So AI chatbots are very overmuch complementing Google searches alternatively than replacing them, moreover though tons of consumers person migrated circumstantial queries distant from the hunt motor successful favour of AI. The champion trading strategies result request to encompass both SEO and GEO.
Shoppers favour AI responses complete AI ads
Until recently, astir of the main AI chatbots person been ad-free experiences. But while companies for illustration OpenAI person been contented to tally astatine ample losses for the liking of personification acquisition, attraction has begun to move to much effective monetization beyond conscionable premium and endeavor plans.
In the US, ads person been progressive successful ChatGPT since February, pinch the beta self-serve advertizing level launching pinch nary minimum walk successful early May. 42.22% of consumers person now seen an advertisement successful an AI chatbot; 35.85% person not seen an ad, while a amazingly precocious 21.94% aren’t judge if they person seen an advertisement aliases not.
Consumer attitudes present are moving fast. When Exploding Topics surveyed consumers successful April, 48.35% expected shopping features and ads to amended AI responses. That study ran while chatbot ads were still successful constricted testing; this 1 ran aft months of wider rollout. Watching sentiment displacement from optimism successful mentation to skepticism successful believe is precisely why brands request to way this abstraction continuously.
The level of uncertainty among consumers arsenic to whether aliases not they person seen a chatbot advertisement speaks to an absorbing point. When a user is utilizing an AI chatbot for merchandise research, the statement betwixt an integrated proposal and a paid advertisement tin quickly go blurred from a user’s standpoint.
Perceived ambiguity betwixt integrated and paid merchandise suggestions helps to explicate why among the 41.63% of consumers who dislike chatbot ads, 66.05% opportunity that ads make them uncertainty the integrity of AI responses. Integrity concerns are by acold the biggest logic for consumers to dislike chatbot ads:
Only astir a 4th (26.81%) of group who dislike chatbot ads specifically opportunity that ads are insufficiently separated from integrated AI content. But moreover erstwhile adverts are intelligibly marked, their specified inclusion is evidently troubling a important number of consumers erstwhile it comes to trust.
Concern astir the integrity of AI chatbot ads climbs pinch age. 68.86% of surveyed users aged 60+ opportunity ads make them uncertainty the integrity of the AI responses, versus 59.13% of surveyed 18-29-year-olds. However, integrity concerns are comfortably the astir communal title astir chatbot ads crossed each property groups.
An rumor from a marketer’s constituent of position is that ads successful AI chatbots person the imaginable to make consumers uncertainty the full response, integrated aliases otherwise. But brands tin only activity connected things wrong their control, and it seems abundantly clear that imaginable customers worth integrated AI mentions much than paid chatbot ads.
The effect of AI chatbot ads
Not each consumers dislike ads successful AI chatbots. 38.09% of group are neutral, and 20.28% for illustration the ads.
We asked the neutral/like group what they for illustration astir ads, and presented them pinch a database of options. More than half (54.85%) said that chatbot adverts acted arsenic an other root of information.
But from a trading standpoint, the much pertinent mobility is whether AI chatbot ads are effective. And our information shows that running AI chatbot ads is perchance much apt to harm your marque than thief it.
20.36% of consumers would person a amended sentiment of a merchandise aft seeing an advertisement for it successful an AI chatbot platform. But for 23.12% of consumers, the advertisement would trigger a worse sentiment of the product.
It’s important to statement that the mostly (56.52%) of consumers said that ads would person a neutral effect connected their sentiment of a product, aliases that it would dangle connected the contented of the ad. So a well-constructed chatbot advertisement still has the imaginable to person a affirmative effect connected consumers. But marketers are efficaciously starting pinch a handicap successful their efforts to execute beardown ROI.
In fairness, 54.04% of group who person seen ads successful chatbots self-report clicking astatine slightest once. But that leaves complete a 3rd (35.85%) who study ne'er clicking, while a further 10.11% are unsure if they person clicked.
Overall, ads successful ChatGPT are besides little apt to lead to a click than ads successful Google SERPs, per self-reported user preferences. 24.88% of consumers opportunity they would beryllium much apt to click connected a ChatGPT advertisement than a Google ad, but 45.42% opportunity they would beryllium much apt to click connected a Google ad.
However, erstwhile you only see consumers who person seen an advertisement successful an AI chatbot, the image flips. 48.94% of consumers who person been served AI ads earlier opportunity they are much apt to click connected a ChatGPT ad, while only 30.64% opportunity they are much apt to click connected a Google ad.
So AI chatbot ads are surely not useless, and clickthrough whitethorn extremity up comparable pinch aliases moreover amended than Google ads arsenic consumers statesman to spot them pinch much frequency. But pinch AI, overmuch for illustration pinch Google SERPs, optimizing to look successful integrated results is apt to nutrient the champion sustainable and semipermanent ROI.
And arsenic a last point, marketers request to see whether their imaginable customers are apt to commencement seeing much AI chatbot ads immoderate clip soon. ChatGPT, for example, only serves ads connected itsand Go plans. Plus, Pro, and business accounts stay ad-free. So for galore B2B products, there’s a consequence that the main target marketplace won’t spot the ads astatine all.
Put GEO astatine the bosom of your AI chatbot strategy
The results we gathered show really consumers are utilizing AI chatbots successful their buying journeys. They are utilizing and trusting AI for marque discovery, merchandise recommendations, and pre-purchase checks, and proposal from a chatbot tin artifact a acquisition arsenic easy arsenic it tin lead to one.
The wider rollout of ads successful starring chatbots for illustration ChatGPT whitethorn look for illustration it presents a elemental avenue into the bottom-of-the-funnel conversations that high-intent users are having pinch AI agents. But advertisers are fighting against a beardown guidelines level of chatbot advertisement skepticism, and tin moreover extremity up doing much harm than bully to a brand’s reputation.
The astir effective and reliable way to harness the ever-growing effect of AI connected merchandise investigation is simply a broad Generative Engine Optimization strategy. GEO is really you make judge consumers are uncovering your marque organically, precisely erstwhile you want them to be.
In practice, that intends 3 things: cognize what AI chatbots presently opportunity erstwhile consumers inquire astir your marque and category, hole the gaps successful the sources those answers tie from, and way your stock of AI recommendations the aforesaid measurement you way rankings.
Semrush provides the champion suite of devices for monitoring and improving your brand’s LLM presence, astatine some the endeavor and SMB scale.
Semrush Enterprise AIO lets you show much than 289 cardinal applicable LLM prompts, search your integrated mentions and uncovering gaps successful your visibility profile. You’ll person instant entree to data-driven recommendations connected really to show up successful much AI queries, encompassing method backend changes and contented fine-tuning.
Your bespoke package will beryllium tailored to your manufacture and goals, and you’ll person entree to broad analytics for benchmarking against competitors. You’ll besides beryllium capable to stock GEO information crossed each of your teams via API.
For smaller businesses, individual marketers aliases freelance SEO specialists and marque strategists, the Semrush AI Visibility Toolkit is the champion GEO instrumentality connected the market. With one-click AI visibility study reports, punctual spread analysis, and continuous share-of-voice tracking, you tin bring each the rigor of elite SEO into the caller era of AI user merchandise research.
More than 80% of AI powerfulness users are already buying products (or not buying them) based connected AI recommendations. Start winning these customers, and book a Semrush Enterprise AIO demo today.
Methodology
We surveyed 2,338 adults successful the US successful July 2026. The separator of correction is positive aliases minus 2 percent points astatine a 95% assurance level for full-sample statistics; filtered segments transportation wider margins.
Respondents were drawn from a operation of regions and economical backgrounds. The median family income fell successful the $75,000-$99,999 band.
The study divided respondents into 4 property bands: 18-29, 30-44, 45-60, and 60+. There were astatine slightest 300 respondents successful each band.
We filtered our information successful various ways successful bid to get the astir broad picture. As such, the denominator sometimes differs betwixt statistic cited successful this report. This is simply a non-exhaustive database of the segments we examined, and really they are referenced successful the report:
- “Consumers”: All study respondents
- “AI users”: Respondents who reported utilizing AI astatine immoderate frequency
- “Regular AI users”: Respondents who reported astatine slightest play AI usage
- “AI powerfulness users”: Respondents who reported utilizing AI aggregate times per week
