AI Shopping and Product Discovery Statistics You Need to Know in 2026

Evidence set

Consumer Adoption of AI Shopping

Consumers are moving from experimentation toward regular use, especially for research, recommendations, and complex purchases. Survey findings measure reported behavior or intent, not completed transactions.

In Adobe's survey, 38% of U.S. consumers had used generative AI for online shopping. (Source: Adobe Digital Insights, Generative AI-Powered Shopping Rises with Traffic to U.S. Retail Sites, 2025)
Adobe found that 52% of surveyed U.S. consumers planned to use generative AI for shopping in 2025. (Source: Adobe Digital Insights, Generative AI-Powered Shopping Rises with Traffic to U.S. Retail Sites, 2025)
Product research was the leading listed AI shopping task, cited by 53% of Adobe survey respondents. (Source: Adobe Digital Insights, Generative AI-Powered Shopping Rises with Traffic to U.S. Retail Sites, 2025)
Among Adobe survey respondents who had used AI for shopping, 85% said it improved their experience. (Source: Adobe Digital Insights, Generative AI-Powered Shopping Rises with Traffic to U.S. Retail Sites, 2025)
Among Adobe's AI-shopping users, 83% said they were more likely to use AI for larger or more complex purchases. (Source: Adobe Digital Insights, Generative AI-Powered Shopping Rises with Traffic to U.S. Retail Sites, 2025)
Salesforce found 63% of Gen Z shoppers wanted AI-agent product recommendations, compared with 23% of baby boomers. (Source: Salesforce, Connected Shoppers Report, 6th Edition, 2025)

AI shopping use and planned adoption

Adobe survey of more than 5,000 U.S. consumers, published in 2025. Used and planned use are separate survey responses. Source: Adobe Digital Insights, Generative AI-Powered Shopping Rises with Traffic to U.S. Retail Sites.

Measure	Value	Scope
Had used AI for shopping	38%	U.S. consumers surveyed by Adobe
Planned to use AI for shopping in 2025	52%	U.S. consumers surveyed by Adobe

How to read this section: The clearest pattern is research-first adoption. People are more comfortable asking AI to compare or explain than handing over the final purchase decision.

Evidence set

How Product Discovery Is Changing

Retail referral data shows rapid growth from a small base, while engagement measures suggest that AI-origin visitors often arrive deeper in the consideration process.

Adobe recorded a 1,300% year-over-year increase in generative-AI referral traffic during the 2024 holiday season. (Source: Adobe Digital Insights, Generative AI-Powered Shopping Rises with Traffic to U.S. Retail Sites, 2025)
Adobe reported a 4,700% year-over-year increase in generative-AI traffic to U.S. retail sites in July 2025. (Source: Adobe Digital Insights, Generative AI-Powered Shopping Rises with Traffic to U.S. Retail Sites, 2025)
Adobe found AI-referred shoppers were 10% more engaged than visitors from non-AI sources. (Source: Adobe Digital Insights, Generative AI-Powered Shopping Rises with Traffic to U.S. Retail Sites, 2025)
AI-referred retail visits lasted 32% longer than visits from non-AI sources in Adobe's comparison. (Source: Adobe Digital Insights, Generative AI-Powered Shopping Rises with Traffic to U.S. Retail Sites, 2025)
Adobe measured a 27% lower bounce rate for retail visits referred by generative-AI sources. (Source: Adobe Digital Insights, Generative AI-Powered Shopping Rises with Traffic to U.S. Retail Sites, 2025)
Google reported that more than one in six U.S. AI Mode queries were entirely non-text in May 2026. (Source: Google, How AI Search Is Reshaping the Growth Playbook for CMOs, 2026)

How AI-referred retail visits differed from non-AI traffic

Adobe comparison of visits to U.S. retail sites by referral source, published in 2025. Each value is a relative difference versus non-AI traffic. Source: Adobe Digital Insights, Generative AI-Powered Shopping Rises with Traffic to U.S. Retail Sites.

Measure	Value	Scope
Higher engagement	10%	Shoppers arriving at U.S. retail sites from generative-AI versus non-AI sources
Longer visits	32%	Visits to U.S. retail sites by source type
Lower bounce rate	27%	Visits to U.S. retail sites by source type

How to read this section: Large growth rates should not be read as market share. Several Adobe comparisons begin from an early, low-volume AI-referral baseline.

Evidence set

Conversational Commerce and Shopping Assistants

Shopping assistants increasingly combine search, recommendations, and comparison in a single conversational flow. The figures below separate stated consumer interest from observed adoption.

In Capgemini's 2024 consumer survey, 71% of consumers said they wanted generative AI integrated into their shopping experience. (Source: Capgemini Research Institute, What Matters to Today's Consumer 2025 press release, 2025)
Capgemini found that 58% of consumers had replaced traditional search engines with generative AI tools for product or service recommendations. (Source: Capgemini Research Institute, What Matters to Today's Consumer 2025 press release, 2025)
Sixty-eight percent of consumers said they want generative AI to combine search-engine, social-media, and retailer results into one set of highlighted purchase options. (Source: Capgemini Research Institute, What Matters to Today's Consumer 2025 press release, 2025)
Forty-six percent of consumers said they were enthusiastic about generative AI's impact on online shopping. (Source: Capgemini Research Institute, What Matters to Today's Consumer 2025 press release, 2025)
Three quarters of consumers, or about 75%, said they were open to recommendations from generative AI. (Source: Capgemini Research Institute, What Matters to Today's Consumer 2025 press release, 2025)
In Adobe's survey, 53% of US consumers said they planned to use generative AI for online shopping during 2025. (Source: Adobe Analytics, Traffic to U.S. Retail Websites from Generative AI Sources Jumps 1,200 Percent, 2025)

What consumers wanted from generative AI in shopping

Capgemini Research Institute survey of 12,000 consumers across 12 countries, conducted in 2024 and published in 2025. Source: Capgemini, What Matters to Today's Consumer 2025.

Measure	Value	Scope
Want AI integrated into shopping	71%	12,000 consumers aged 18+ across 12 countries
Replaced traditional search for recommendations	58%	12,000 consumers aged 18+ across 12 countries
Want combined search, social, and retailer results	68%	12,000 consumers aged 18+ across 12 countries
Enthusiastic about AI shopping impact	46%	12,000 consumers aged 18+ across 12 countries
Open to generative AI recommendations	75%	12,000 consumers aged 18+ across 12 countries

How to read this section: Preference and planned-use surveys signal demand, but they do not prove revenue, conversion, or sustained use by themselves.

Evidence set

Recommendations and Personalization

Consumers use AI to compare products and generate inspiration, while retailers use AI to improve recommendations and cross-sell opportunities. Channel comparisons remain observational rather than causal.

McKinsey's France-Germany-UK survey found that 63% of consumers use AI to compare brands, models, prices, and reviews. (Source: McKinsey, Europe's agentic commerce moment: Decision influence is here; execution is coming, 2026)
Fifty-five percent of consumers in McKinsey's France-Germany-UK survey used AI to learn about a product category or product. (Source: McKinsey, Europe's agentic commerce moment: Decision influence is here; execution is coming, 2026)
Forty-six percent of consumers in McKinsey's France-Germany-UK survey used AI to discover products or get purchase inspiration. (Source: McKinsey, Europe's agentic commerce moment: Decision influence is here; execution is coming, 2026)
Adobe Analytics found that visitors arriving at US retail sites from generative-AI sources had 8% higher engagement than visitors from non-AI sources. (Source: Adobe Analytics, Traffic to U.S. Retail Websites from Generative AI Sources Jumps 1,200 Percent, 2025)
Adobe Analytics found that retail visits from generative-AI sources were 9% less likely to convert than visits from other traffic sources. (Source: Adobe Analytics, Traffic to U.S. Retail Websites from Generative AI Sources Jumps 1,200 Percent, 2025)
Adobe's 2025 retail report found that 41% of retail organizations were using AI for smarter product recommendations aimed at cross-sell and upsell opportunities. (Source: Adobe and Econsultancy, 2025 AI and Digital Trends for Retail, 2025)

How European consumers use AI before purchase

McKinsey survey of consumers in France, Germany, and the United Kingdom, published in 2026. Source: McKinsey, Europe's agentic commerce moment.

Measure	Value	Scope
Compare brands, models, prices, and reviews	63%	Consumers in France, Germany, and the United Kingdom
Learn about a category or product	55%	Consumers in France, Germany, and the United Kingdom
Discover products or get inspiration	46%	Consumers in France, Germany, and the United Kingdom

How to read this section: Higher engagement can coexist with lower conversion when shoppers use AI earlier in the journey. The metric and the stage of the funnel both matter.

Evidence set

Merchant Adoption and Reported Outcomes

Retailers report broad AI adoption, expanding budgets, and positive commercial effects. Most figures in this section are survey responses, so they reflect reported experience rather than audited financial statements.

In NVIDIA's 2025 retail and CPG survey, 89% of respondents were already using AI or assessing projects such as trials and pilots. (Source: NVIDIA, State of AI in Retail and CPG: 2025 Trends survey summary, 2025)
NVIDIA's 2025 retail and CPG survey found 87% of respondents reporting a positive impact from AI on annual revenue. (Source: NVIDIA, State of AI in Retail and CPG: 2025 Trends survey summary, 2025)
Ninety-four percent of respondents in NVIDIA's 2025 retail and CPG survey said AI helped reduce their annual operating costs. (Source: NVIDIA, State of AI in Retail and CPG: 2025 Trends survey summary, 2025)
Forty percent of respondents in NVIDIA's 2025 retail and CPG survey cited digital shopping assistants or copilots as a generative-AI use case. (Source: NVIDIA, State of AI in Retail and CPG: 2025 Trends survey summary, 2025)
Salesforce's sixth Connected Shoppers report says 84% of surveyed retailers used AI, while just 2% had no plans to use it. (Source: Salesforce, Connected Shoppers Report, 6th Edition, 2025)
Among retailers using AI, 89% in Salesforce's report expected to get a return on their AI investments. (Source: Salesforce, Connected Shoppers Report, 6th Edition, 2025)

Retail and CPG professionals reported broad AI use and benefits

NVIDIA State of AI in Retail and CPG survey summary, published in 2025. Findings are respondent-reported outcomes, not audited financial statements. Source: NVIDIA, State of AI in Retail and CPG: 2025 Trends.

Measure	Value	Scope
Using or assessing AI	89%	Retail and CPG industry professionals responding to NVIDIA's survey (hundreds of responses)
Reported positive annual revenue impact	87%	Retail and CPG industry professionals responding to NVIDIA's survey
Reported lower annual operating costs	94%	Retail and CPG industry professionals responding to NVIDIA's survey

How to read this section: Industry evidence can establish context, but it cannot substitute for a brand's own attribution, conversion, and revenue records.

Evidence set

AI Referral Traffic and Commerce Performance

Observed retail analytics show that AI-origin traffic is growing quickly and that its conversion profile changed materially between 2025 and 2026.

Adobe Analytics recorded a 693.4% year-over-year increase in generative-AI traffic to retail sites during the 2025 holiday season. (Source: Adobe Digital Insights, AI traffic surges across industries, retail sees biggest gains, 2026)
During the 2025 holiday season, Adobe found retail visits referred by AI converted 31% more than visits from other traffic sources. (Source: Adobe Digital Insights, AI traffic surges across industries, retail sees biggest gains, 2026)
Adobe reported a 254% year-to-date increase in revenue per visit from AI-driven retail traffic during the 2025 holiday period. (Source: Adobe Digital Insights, AI traffic surges across industries, retail sees biggest gains, 2026)
Adobe measured a 393% year-over-year increase in AI-source traffic to U.S. retail sites in January-March 2026. (Source: Adobe Digital Insights, AI traffic grows but retail sites lag in AI search visibility, 2026)
Adobe's March 2026 retail data showed AI-referred traffic converting 42% better than non-AI traffic. (Source: Adobe Digital Insights, AI traffic grows but retail sites lag in AI search visibility, 2026)
Adobe Digital Insights measured a 35-times increase in AI-driven retail referrals over the 11 months from July 2024 to May 2025. (Source: Adobe Digital Insights, Q2 2025 insights: AI referrals surge across industries, 2025)

How to read this section: These are channel comparisons, not controlled experiments. They show association with traffic quality and conversion, not proof that AI caused the outcome.

Evidence set

Trust, Accuracy, and Shopping Friction

Adoption depends on accuracy, transparency, data control, and the ability to reverse or verify an AI agent's work. Consumers still want meaningful oversight.

Gartner found that 54% of U.S. consumers who used AI for a recent purchase had to double-check every piece of information the GenAI tool supplied. (Source: Gartner, Gartner Survey Finds Consumers Want AI Shopping Help, But Not AI Purchase Decisions, 2026)
In Gartner's survey, 62% of U.S. consumers who used AI during a recent purchase said the information ultimately wasted their time. (Source: Gartner, Gartner Survey Finds Consumers Want AI Shopping Help, But Not AI Purchase Decisions, 2026)
Visa's three-market consumer research found that nearly nine in ten respondents want to see how an AI shopping agent makes decisions. (Source: Visa, Earning Consumer Trust in the Age of Agentic Commerce, 2025)
Visa reports that about half of consumers would stop using an AI shopping agent if they lost control over how it operated. (Source: Visa, Earning Consumer Trust in the Age of Agentic Commerce, 2025)
Capgemini's 2026 consumer research found that 71% of consumers worry about how generative AI uses their data. (Source: Capgemini Research Institute, What Matters to Today's Consumer 2026 infographic, 2026)
Two-thirds of consumers told Capgemini that brands should clearly label or otherwise inform them when an ad is AI-generated. (Source: Capgemini Research Institute, What Matters to Today's Consumer 2026 infographic, 2026)

Recent AI-shopping users still faced verification costs

Gartner survey of U.S. consumers who used AI while shopping for a recent purchase, published in 2026. Source: Gartner, Consumers Want AI Shopping Help, But Not AI Purchase Decisions.

Measure	Value	Scope
Checked every AI result	54%	U.S. consumers who used AI while shopping for a recent purchase
Said AI information wasted time	62%	U.S. consumers who used AI while shopping for a recent purchase

How to read this section: Trust is operational. Shoppers want to understand recommendations, see what data is collected, and retain control over consequential decisions.

Evidence set

Market Outlook for AI-Influenced Commerce

Forecasts point toward more AI-influenced orders, while post-season data shows that AI assistants already affected discovery and engagement during the 2025 holiday period.

Salesforce's September 2025 forecast put AI- and agent-referred activity at 21% of global holiday orders, worth an estimated $263 billion. (Source: Salesforce, With $1.25T in Holiday Sales Up for Grabs, LLMs Emerge as New Frontier for Search, 2025)
Salesforce recorded 119% year-over-year growth in online traffic from AI assistants during the first half of 2025. (Source: Salesforce, With $1.25T in Holiday Sales Up for Grabs, LLMs Emerge as New Frontier for Search, 2025)
Salesforce reported that AI-assistant referrals converted at more than 700% above social-media referral rates in the first half of 2025. (Source: Salesforce, With $1.25T in Holiday Sales Up for Grabs, LLMs Emerge as New Frontier for Search, 2025)
Salesforce's observed Nov. 1-Dec. 31, 2025 holiday data attributes 20% of all retail sales, or $262 billion, to AI and agents through recommendations and engagement. (Source: Salesforce, Holiday Season Rakes in Record $1.29T for Retailers, Salesforce Data Shows, 2026)
Visa's global holiday study found that 47% of U.S. consumers had used AI for at least one shopping task, including gift discovery, price comparison or product research. (Source: Visa and Morning Consult, AI and Digital Wallets Reshape Spending Habits, 2025)
Visa's December 2025 outlook predicted that millions of consumers would let AI agents complete purchases by the 2026 holiday season. (Source: Visa, Visa and Partners Complete Secure AI Transactions, Setting the Stage for Mainstream Adoption in 2026, 2025)

How to read this section: Forecasts and observed results are labeled separately. A projection is not treated as a completed sale or verified outcome.

Quality controls

Methodology and Limitations

This page was built from a pool of 81 candidate facts. Six duplicate records and one misread source value were removed, leaving 74 unique valid candidates. Forty-eight were selected for balance, recency, source diversity, and usefulness to reporters.

Every retained source URL was fetched and recorded in the research audit.
All 48 displayed statistics come from official primary sources dated 2025 or 2026.
Survey intention, observed behavior, reported outcomes, and forecasts are labeled as different evidence types.
Relative traffic growth is not presented as market share, and channel comparisons are not presented as causal proof.
No Teal-specific merchant outcome appears without qualifying first-party evidence.

What this evidence cannot prove

Survey answers describe reported behavior or intent and do not establish completed purchases.
Referral-traffic comparisons are observational and do not establish that AI caused a conversion outcome.
Large year-over-year growth rates can begin from a small base and should not be read as market share.
Retailer-reported revenue and cost effects are not substitutes for audited financial statements.
Forecasts describe modeled future outcomes, not observed transactions.

Quick answers

Frequently Asked Questions

How many people use AI for online shopping?

The answer depends on market, date, and definition. Adobe reported that 38% of surveyed U.S. consumers had used generative AI for online shopping in 2025, while other studies measured planned use, familiarity, or specific shopping tasks rather than completed purchases.

What do shoppers use generative AI for?

The most common uses in the cited research include product research, comparisons, recommendations, deal discovery, and help with larger or more complex purchases. These activities usually happen before checkout.

Does AI referral traffic convert better than other traffic?

The relationship changed over time. Adobe reported AI-referred retail traffic lagging non-AI conversion in parts of 2025, then outperforming other sources during the 2025 holiday season and again in March 2026. These are observed channel comparisons, not causal experiments.

What prevents consumers from trusting AI shopping agents?

Accuracy checks, unclear data use, loss of control, and opaque recommendations are recurring barriers. Gartner and Visa research both show that shoppers want assistance without surrendering final oversight.

Are these statistics forecasts or measured results?

Both appear, and they are labeled. Survey intentions, forecasts, observed web analytics, and reported merchant outcomes are different evidence types and should not be combined as if they measured the same thing.

How were the sources checked?

Each retained statistic was traced to a fetched official source page or report, recorded with its URL, year, evidence location, method note, and qualification in the companion research audit. Each displayed claim links to that official source, and the bibliography lists each source document once.

Primary research

Sources

Each source below links to the exact official page or report used in the research audit. Claim-level citations above point to the same documents.