77% use AI to shop. Nearly 1 in 3 won’t let it spend.

Half 1: The AI Commerce Surge

AI is an more and more ubiquitous buying instrument

To get a baseline, we first requested respondents how they might describe their use of AI usually. This yielded a placing response. 

Of the 1,000+ individuals we surveyed, virtually half reported utilizing AI regularly. However extra eye-catchingly, solely 9.81% had by no means used an AI instrument.

Excessive adoption was a recurring theme after we requested about AI commerce particularly. 43.21% of shoppers are utilizing AI to assist with buying not less than as soon as per week, with effectively over half of buyers utilizing the know-how not less than month-to-month.

Amongst “frequent” AI customers, solely 2.76% had by no means used the know-how to assist with buying or buying choices. That determine rose to 22.4% general, leaving over three-quarters of shoppers who’ve not less than tried utilizing AI for buying.

Even among the many over-60s, greater than half (52.78%) reported utilizing AI for buying within the final six months. 18.75% of them use AI buying instruments weekly or extra.

We put the following set of questions completely to the 77.6% of respondents who have adopted AI commerce, with a purpose to higher perceive how they’re utilizing AI to buy.

How are shoppers utilizing AI to buy?

“AI commerce” is a broad time period that may seize a variety of shopper actions. We wished to search out out precisely how individuals are incorporating AI into their purchases.

Product analysis emerged because the main use case, adopted by greater than two-thirds (68.5%) of buyers. Greater than half (55.19%) additionally reported utilizing AI for locating the very best value/offers.

Deciding between manufacturers, getting present concepts, and summarizing buyer evaluations have been all additionally moderately frequent (>353.3%).

Apparently, buyers even have clear preferences about what they are going to purchase with the assistance of AI. Customers are almost definitely to make use of AI help when looking for clothes or know-how.

This knowledge actually underlines how embedded AI has change into in buying routines already. 44.62% have used it for one thing as mundane as grocery buying. 

Extra “distinctive” and doubtlessly costly purchases, the place you would possibly intuitively assume that customers would flip to AI for a bit of additional help, are usually much less frequent use circumstances. Furnishings (29.62%) and jewellery (28.08%) have been among the many least fashionable responses.

The AI buying instruments of alternative

ChatGPT nonetheless enjoys the biggest market share throughout most shopper AI capabilities, and buying isn’t any totally different. 77.56% of buyers use ChatGPT when they need AI help.

The prevalence of Gemini is probably extra shocking. 58.21% of buyers reported utilizing Google’s AI, effectively over twice as many as the following hottest instrument.

In fact, Google has embedded Gemini in AI Overviews and AI Mode. Its search engine has lengthy been the go-to instrument for handbook product analysis, so maybe it ought to be no nice shock that its AI has now captured a whole lot of the identical market.

However intriguingly, Gemini utilization is definitely most saturated amongst individuals who use AI to put in writing buying lists. 75.86% of AI list-writers report utilizing Google’s instrument to buy (in comparison with 60.49% of those that use AI for product analysis).

That’s to not say Gemini is essentially getting used to put in writing these buying lists. However the anticipated skew towards Gemini amongst those that reported utilizing AI for product analysis merely didn’t materialize, suggesting that buyers could effectively have a real desire for Google’s AI past simply its embedded search options.

Fewer than one in 5 individuals utilizing Claude for AI commerce was additionally notable. Anthropic’s instrument truly overtook ChatGPT in an enterprise context final yr, however adoption for on a regular basis shopper duties remains to be decrease than opponents. 

Grok stays essentially the most extremely gendered instrument. It’s utilized by 31.98% of male buyers, however simply 15.16% of ladies.

Throughout the board, males have been extra doubtless than girls to make use of AI instruments for buying. Nonetheless, ChatGPT utilization was near equal (78.05% of males vs 77.51% of ladies).

Evolving AI buying habits

It’s outstanding how rapidly AI has embedded itself as a regular buying companion. Amongst those that are actually utilizing the know-how, 39.1% say they use AI for buying “rather more” than they did six months in the past.

An extra 28.97% of shoppers are utilizing AI instruments for buying “a bit extra” within the final half-year. Solely 6.02% have decreased their utilization.

Center Atlantic residents stand out because the keenest adopters. Virtually half (49.04%) are utilizing AI for buying rather more prior to now six months, and near eight in 10 (78.98%) have not less than considerably elevated their utilization. West North Central is the least enamored with the know-how, with over 13% utilizing AI for buying much less regularly than they did beforehand.

Nationwide, the affect of the know-how on buying habits is stark. 92.54% of shoppers say it’s not less than potential AI has immediately influenced them to purchase one thing they wouldn’t have in any other case bought.

Virtually seven in 10 (68.64%) can positively keep in mind being immediately influenced to make a purchase order. That features 36.89% who say they’ve been influenced “many occasions” by AI.

This development is most pronounced among the many highest earners. 61.9% of shoppers with a family earnings of $125,000 or increased have made AI-influenced purchases “many occasions,” and solely 13.19% can not recall any such buy.

Why the elevated uptake?

Though the pace of AI buying adoption is startling, the explanations behind it are in the end no thriller. Fairly merely, the bulk of people that have tried utilizing AI instruments have discovered that they make product analysis simpler.

37.18% say that AI makes buying analysis a lot simpler. An extra 40.9% say AI makes it considerably simpler. 

For essentially the most half, shoppers additionally belief AI as a buying instrument. 

Solely round one in 5 buyers say that they belief AI utterly. However that rises above 60% when additionally counting those that principally belief AI as a buying instrument, with some handbook fact-checking.

In some ways, that is the anticipated sample, provided that the query was solely put to individuals who have tried utilizing AI as a buying instrument. These with the least belief could not have tried it within the first place.

Nonetheless, it’s fairly a pointy distinction from one other of our unique surveys, assessing attitudes to AI Overviews. In that context, 82% of respondents were at least somewhat skeptical of the outputs, and but the overwhelming majority continued to depend on them anyway (with out routinely checking sources) for the sake of comfort.

Relating to buying, customers appear to have extra real religion in AI outputs: They’re utilizing it not just for its comfort, however as a result of it usually works effectively. That may very well be an indication of common AI enhancements within the ~9 months between the surveys, or it might be an indication that commerce is an space the place the know-how can actually excel for shoppers.

The standard AI buy pipeline

So most individuals are utilizing AI commerce instruments, and uptake has solely gotten increased within the final six months. However apparently, there isn’t any clear consensus about how to make use of AI for buying.

We all know that product analysis and value comparability is fashionable. However that doesn’t inform us an excessive amount of about what a typical AI-assisted buying journey truly appears like.

We gave respondents 4 choices:

• I exploit AI as a place to begin after which seek the advice of different sources
• I begin on conventional retail web sites after which use AI as a complement
• I exploit AI as my solely supply after which full checkout externally
• I full your complete buying course of in AI, from preliminary analysis to checkout

There was an virtually precisely even cut up between the primary two choices. 44.8% begin on retail web sites after which add in AI, whereas 44.03% use AI as a place to begin earlier than trying externally.

That is notable for retailers, and underlines the paramount significance of Generative Engine Optimization (GEO). 

An enormous base of potential prospects are utilizing AI as a place to begin, so it’s crucial that your brand gets organically mentioned. And for these beginning in your web site however then double-checking with AI, brand sentiment might make or break a sale.

The opposite factor that stands out from this knowledge is that utilizing AI for your complete buying journey remains to be a fringe use case. Solely 8.99% of customers are utilizing AI as their solely supply earlier than buying, and solely 2.18% are trying out by way of AI. 

In Half 2, we’ll look at the the reason why. Questions within the second half have been put to all respondents, to get a greater concept of the present attitudes held by each adopters and non-adopters of AI commerce. 

Spot the following “AI commerce” 12 months early. Exploding Matters Professional tracks 11M+ developments with search quantity, progress curves, and class filters. Start your 7-day free trial. 

Half 2: The AI commerce crimson line

On the spot Checkout: Don’t understand it, don’t prefer it

No matter which stage within the buying course of customers introduce synthetic intelligence, the ultimate step is sort of at all times exterior checkout. Given that customers are clearly eager on utilizing AI as a part of the commerce journey, instruments that remove this level of friction make superficial sense. 

That was the concept behind Instant Checkout from ChatGPT; you are able to do your whole analysis throughout the app, after which full your buy there as effectively. In impact, the AI agent completes the transaction in your behalf. 

OpenAI isn’t the one one to construct one thing like this. Visa’s Intelligent Commerce is an identical payment-side resolution, whereas Google has developed its own AP2 protocol for “agent-led funds.” 

However consciousness of latest and upcoming instruments that can help you checkout from immediately inside an AI interface is kind of low. 42.83% of individuals have been by no means conscious, with an extra 23.01% solely “vaguely conscious.” 

Unsurprisingly, those that use AI for buying weekly or extra are almost definitely to be “very conscious” of On the spot Checkout and comparable instruments (63.3%). However that drops to 25.19% amongst month-to-month customers, and simply 11.11% amongst those that have used AI buying instruments “a couple of occasions.” 

Having been instructed concerning the existence of those instruments, the response might greatest be described as combined. 

From a preset checklist of choices, “skeptical” was chosen most frequently (41.083%), adopted by “suspicious” (33.1%). However respondents might choose a couple of reply, and “excited” (31.61%), “completely satisfied” (24.33%), and “impressed” (24.03%) have been the following most-common solutions.

Those that chosen to fill in a solution of their very own have been overwhelmingly damaging. Responses included “hunted/preyed upon,” “terrified,” “cautious,” and “not .” 

Crucially, there was important negativity towards On the spot Checkout even amongst those that are already routinely utilizing AI instruments for buying.

29.82% of essentially the most common AI buying customers mentioned they have been suspicious of instruments like On the spot Checkout, and 29.59% reported being skeptical. Amongst month-to-month customers, skepticism was the only hottest angle (37.04%). 

In the meantime, solely 2.22% of the individuals who aren’t at the moment utilizing AI to buy reported being excited on the prospect of brokers having the ability to perform buy orders. 

In reality, the concept of AI buying energy is actively making non-users much less more likely to attempt AI for buying. 

44.89% of AI buying non-adopters are “a lot much less doubtless” to attempt the know-how on account of these new instruments. Over half are not less than a bit much less doubtless, and solely 7.11% are extra doubtless. 

Then again, essentially the most common present AI buyers anticipate that instruments like On the spot Checkout will additional improve their utilization. 72.71% say that the improvements make them not less than considerably extra more likely to store with AI extra often.

Exterior of energy customers and non-users, indifference is extra frequent. 48.89% of month-to-month customers anticipate On the spot Checkout (and comparable instruments) will make no distinction to their utilization, as do 52.17% of occasional customers.

And it appears OpenAI will need to have reached an identical conclusion. Mere months after launching On the spot Checkout, it has rowed back on direct shopping features, doubling down on the invention aspect of issues.

Mistrust of AI firms with fee knowledge

One of many largest hurdles on the subject of additional integrating AI into commerce is that most individuals don’t really feel snug trusting chatbots with their card particulars with a purpose to make direct purchases simpler in future.

In complete, 51.45% of shoppers are not less than considerably uncomfortable on the concept of AI instruments storing their card particulars. Solely round 1 in 4 are “very snug.” 

In addition to being the preferred response general, “very uncomfortable” additionally lower throughout age teams to an sudden diploma. Greater than a 3rd of shoppers aged 18-29 mentioned they might be very uncomfortable storing card particulars with an AI instrument, regardless of being digital fee natives.

Even among the many most frequent AI buyers, barely greater than half (50.69%) mentioned they might be “very snug” with AI instruments storing their card particulars. That dropped dramatically to 18.52% amongst month-to-month AI buyers, 7.25% amongst those that use the know-how sometimes, and simply 0.89% amongst those that don’t at the moment use AI to buy in any respect.

Pacific residents are almost definitely to belief AI instruments with their card particulars, with 64.48% not less than considerably snug, whereas the Center Atlantic as soon as once more stands out as a distinctly pro-AI area. New England is essentially the most distrustful (58.53% not less than considerably uncomfortable).

Who does AI commerce serve?

Tied in with this discomfort about fee particulars is the truth that shoppers are skeptical of whether or not they’re really the supposed beneficiaries of AI commerce know-how.

Solely 14.16% of respondents mentioned shoppers are those being primarily served by AI buying instruments proper now.

The most typical reply (27.52%) was that these instruments are made to serve the pursuits of AI firms themselves. Manufacturers and advertisers (27.32%) was one other fashionable response.

And even among the many most frequent customers of AI buying instruments, solely 23.85% of shoppers imagine they’re those whom the instruments are primarily serving. These energy customers have been extra more likely to say that manufacturers and advertisers are those being served.

Amongst much less frequent customers, skepticism rises sharply, to the purpose the place simply 2.22% of non-users imagine AI buying instruments are primarily serving shoppers proper now.

“The mode quantity a shopper would authorize AI to spend autonomously is $0.”
— Exploding Topics, 2026 shopper AI commerce survey

Laborious spending cap for autonomous AI purchases

Provided that a point of skepticism cuts throughout a number of demographics, it isn’t too shocking to study that customers stay reluctant to empower AI to spend huge sums autonomously. 

Nonetheless, the extent of the reluctance is eye-catching: the mode quantity a shopper would authorize AI to spend autonomously is $0.

Particularly, we requested how a lot shoppers would belief AI to spend within the state of affairs the place they have been instructing an AI agent to purchase one thing as soon as it turned obtainable. This hypothetical aligns intently with the acknowledged use circumstances of the most recent AI commerce improvements, together with Google’s AP2 Protocol:

However proper now, our survey reveals the urge for food is just not there. 31.21% of shoppers wouldn’t permit any autonomous AI spend in any respect, 17.45% would cap it at $20, and 20.74% would cap it at $50.

This instantly all however wipes out one other of Google’s proposed use circumstances: the instance of instructing an AI to purchase live performance tickets the second they go on sale. Assuming most such transactions would exceed $100 complete, solely 11.71% of shoppers would at the moment be snug trusting AI with the acquisition.

AI firms even face a tough promote amongst their common customers. 51.84% of weekly AI buyers would cap autonomous AI spend at $50 or much less, as would 67.41% of month-to-month buyers.

Barely a couple of in 5 (20.87%) of essentially the most frequent AI buying customers could be ready to authorize a spend over $100.

Unsurprisingly, the very best earners are the almost definitely to belief AI to extend purchases. However even then, 68.57% would cap brokers at $100 or much less: 1/2000th of their annual family earnings at most.

Agentic commerce is right here to remain

The strain on the coronary heart of those outcomes is that regardless of this reluctance to sanction AI spend, there may be widespread perception that AI’s position in commerce will proceed to get greater.

Greater than half of individuals (55.83%) assume AI will play an even bigger position in how they store in 5 years’ time. Solely 12.37% imagine it’s going to play a smaller position.

Even amongst non-users, virtually a 3rd (32.44%) predict that AI will play an not less than considerably greater position in how they store in 5 years’ time. And 74.77% of essentially the most frequent AI buyers imagine the know-how will tackle a fair greater position in how they make purchases.

A way forward for expanded AI commerce would include additional questions. As an example, a landscape of ads and sponsored links has the potential to disrupt the standard of AI outputs.

Nonetheless, most shoppers appear glad that elevated AI buying options gained’t actively affect the standard of responses. In reality, 48.35% imagine the rollout of extra buying capabilities and the mixing of adverts will truly enhance the general commonplace of AI solutions.

Solely round one in 10 of essentially the most frequent customers predict that adverts and buying options will make AI outputs worse, a discovering which AI firms might effectively interpret as one thing of a inexperienced gentle to push forward with this type of monetization.

The ultimate certain signal that AI commerce will proceed to develop is just that buyers prefer it. Even when there may be some skepticism about whether or not shoppers are really the principle beneficiaries, 55.83% agree that AI options make buying not less than considerably higher for shoppers general.

Amongst customers and non-users alike, fewer than one in 5 individuals assume AI has made the buying expertise any worse. That falls under one in 10 amongst those that have used the know-how in any respect throughout the final six months.

Although there are disagreements about what path this burgeoning know-how ought to take subsequent, it appears more and more clear that AI buying is right here to remain.

For those who’re a retailer, a instrument like Semrush Enterprise AIO is extra essential than ever. 77% of your prospects are utilizing AI of their commerce journeys, and the visibility and popularity of your model has the potential to remodel your backside line.

Request a live demo today.

Methodology

This survey was accomplished by 1,009 respondents in complete. After the overall questions on frequency of AI utilization and AI buying utilization, non-users have been skipped for the rest of Half 1, earlier than being reintroduced for Half 2 (the place the opinions of non-users provided useful insights).

All respondents have been from the US, spanning all areas (there have been no respondents from the US Territories). 56.03% of respondents have been feminine, and 43.97% have been male. 

10 totally different family earnings bands have been represented, from $0-9,999 as much as $200,000+. The median earnings vary was $75,000-$99,999.

The age vary was as follows: 

• 14.76% aged 18-29
• 43.47% aged 30-44
• 27.42% aged 45-60
• 14.36% aged over 60

Want the key stats in a one-pager? Download the full summary.