AI Is Winning Over Shoppers: New Report Finds 45% Don’t Care if Recommendations Come from Humans or Algorithms
Constructor and Shopify have released their third annual State of Ecommerce report, and the findings capture how far artificial intelligence (AI) has moved from novelty to habit. Based on a survey of 1,500 consumers across the U.S., U.K., and Germany, the 2025 study shows that nearly half of shoppers (45%) are indifferent to whether product picks come from people or algorithms. As long as the recommendation fits, the source doesn’t matter.
AI Moves From Hype to Habit
Generative AI (GenAI) tools are no longer confined to side projects or experiments. Two-thirds of shoppers (64%) report using GenAI in daily life, compared to 51% last year and 29% in 2023. In retail, nearly 6 in 10 consumers are now comfortable using conversational AI on shopping sites, up from 42% in 2023. This signals clear normalization of AI as part of how people shop, not a future concept waiting for adoption.
AI Agents Step Into the Aisle
Nearly 8 in 10 shoppers arrive at retail websites unsure of what to buy. Gift shopping, new hobbies, and event outfits all drive uncertainty. In these moments, 60% of consumers are open to letting an AI agent provide guidance. Agentic shopping tools like Amazon’s Rufus and Walmart’s Sparky are already in circulation, with 38% of shoppers having tried them. First impressions are positive: half of users say these features are always or often helpful.
Trust and Taste
Shoppers are giving algorithms a vote of confidence on personal style. More than a quarter (26%) say an algorithm that uses browsing and purchase history is more likely to get their taste right than an influencer (18%). Nearly one in five (19%) trust an AI agent more than their partner to pick out a gift. Generational splits are sharp: 25% of Gen Z pick AI over a partner for gifts, compared to 7% of Boomers. It reads like a punchline, yet it points to growing reliance on machine-driven picks for choices that once felt personal.
Discovery Channels by Generation
TikTok and Google Lead
Discovery paths vary by age. Nearly half of Gen Z begin searches on TikTok. Across all ages, 84% still start with Google. Amazon remains a frequent launch point at 63%, with regional differences: 75% in the U.S., 63% in Germany, 53% in the U.K. Only 23% start on a retailer’s site, showing how tough it is to win the first click.
Social Platforms Shape Younger Cohorts
TikTok and Instagram drive discovery for Gen Z and millennials. Facebook stays strong for Gen X. Many Boomers skip social shopping entirely. App and device use mirror this split: half of Gen Z prefer retailer apps; most Boomers lean to desktop sites.
The Search Problem That Won’t Go Away
Two-thirds of shoppers (68%) think on-site search needs an upgrade. The pain is familiar: 42% often reformulate queries; 41% feel treated like strangers despite past activity; 50% see results that are technically on-topic yet not attractive. Nearly half spend three minutes or more plowing through results, and 23% spend over eight minutes. Many rank this grind as worse than waiting at the DMV, sitting in traffic with no AC, or a bad blind date.
The Cost of Bad Search
Weak discovery pushes people away. Forty-seven percent leave a site immediately when search disappoints. Forty-four percent head to another retailer. Two-thirds switch to Amazon after a failed search. On the flip side, more than half would pay 5–10% more if they could skip endless scrolling. Fixing search does not just remove friction; it creates value.
Retail Media: Mixed Sentiment
Retail media spend is rising, but patience is thin. A third of shoppers click sponsored listings, yet nearly half are skeptical. Over a quarter call these ads annoying. Common complaints: better results pushed down the page, weak relevance, and clutter. Only 6% say they like sponsored placements. Ads must meet the same standard as organic results or they read as noise.
Historical Trendlines
Adoption moved fast over three years: GenAI use climbed from 29% in 2023 to 64% in 2025. Comfort with on-site conversational tools rose from 42% in 2023 to 58% in 2025. The story is steady growth, not a one-off spike. Attitudes toward AI agents also shifted from novelty to practical help, as trial rates reached 38% and helpfulness ratings hit 50% among users.
Industry Benchmarks and Context
Shoppers grade every site against leaders. One-click wallets, clean product data, and clear shipping dates have set a high bar. When a retailer’s search cannot parse “black hoodie large,” people bounce to Google or Amazon. Industry UX studies and checkout audits echo this pattern: intent handling and data quality drive conversion more than flashy features.
The Revenue Impact of Broken Discovery
Percentages hide the real hit. Take a brand at $10,000,000 in monthly online sales. If weak search drives quick exits for 47% of sessions, rescuing even one point of conversion can add six figures a year. Every extra field at checkout, every vague delivery date, and every irrelevant recommendation carries a price tag.
Cultural Notes on Trust and Taste
AI beating partners on gift picks grabs attention, yet the deeper point is preference for outcomes over origins. People want the right item at the right time. If the pick feels right, they do not dwell on who—or what—picked it. That pragmatism explains rising comfort with agentic tools and the slow fade of influencer sway for some categories.
What Retailers Should Do Now
Fix Search Relevance
- Support natural language queries and common synonyms.
- Bias results to in-stock, well-rated, high-margin items with clear disclosure.
- Show query understanding on the page: “Searching for black hoodie in large.”
Clean Product Data
- Standardize titles, attributes, sizes, and colors across the catalog.
- Enrich listings with specs, sizing help, and care details drawn from source data.
- Suppress items already purchased unless it is a clear replenishment.
Streamline Checkout
- Cut fields and steps; offer wallets and pay-over-time options.
- Surface taxes, fees, and returns up front to prevent last-second exits.
- Auto-save carts and keep them consistent across devices.
Set Delivery Expectations
- Show delivery windows and cut-off times on product pages, not just at checkout.
- Offer store pickup, lockers, and reliable next-day options where possible.
- Provide live order status without forcing account creation.
Smarter Retail Media
- Hold sponsored placements to the same relevance standards as organic results.
- Limit ad density; avoid pushing the best organic results below the fold.
- Label ads clearly and explain why the item appears.
Build Trust in AI
- Offer opt-in AI help that explains sources and respects privacy.
- Summarize reviews into pros and cons; answer PDP FAQs from verified data.
- Never let AI hijack the cart; keep the shopper in control.
Measure What Matters
- Track search-to-add-to-cart, not just clicks.
- Segment abandonment by step to find the real blocker.
- Use holdout tests to prove personalization lift.
My Take on the Data
I’ve run enough e-commerce audits to see the same pattern on repeat. Teams chase shiny features and skip the basics that move revenue. Shoppers do not care about sizzle if size filters break or delivery dates hide behind a login. AI will help the sites that get the basics right. For the rest, AI becomes one more layer on top of weak foundations.
The 2025 State of Ecommerce report points to a clear path. AI is routine, search is still a headache, and shoppers pick outcomes over origins. Brands that fix data quality, sharpen search, clarify delivery, and use AI as a helpful guide will win repeat business. Brands that do not will keep funding their rivals, one abandoned session at a time.