Consumers are still concerned about privacy, but they are getting more comfortable trading data for better product recommendations. Share a little about what you’ve browsed or bought, and the experience becomes faster and more relevant.
But that openness has limits, and when brands go beyond what feels right or transparent, trust quickly erodes.
To improve recommendations, 43% of American consumers say they will share their browsing history, 42% their past purchase history and 34% their location, according to Omnisend’s “AI Shopping Report.” That’s a reasonable level of openness, but it comes with clear expectations about data usage.
AI-driven shopping is a direct exchange. Consumers share data to get better recommendations, faster decisions and less friction.
What data can you share to improve recommendations?
Compatibility drives data sharing, but not without limitations
People will share data if the payment is visible. Behavioral indicators like browsing and purchase history feel welcome because they directly improve the experience, and accurate recommendations make them willing to share more.
Your customers are searching everywhere. Make sure it’s your product he appears.
The SEO toolkit you know, and the AI visibility data you need.
Start a Free Trial
Start with
That change increases consumers’ faith in AI search results. According to Omnisend, 42% of US consumers say ChatGPT provides better product recommendations than traditional search engines, indicating a shift away from search to recommendation-driven discovery.
But commitment declines when the connection to value is unclear. Public data still feels personal, and consumers are less comfortable sharing it for personalization.
Trust breaks down in certain areas
Relying on AI procurement is not an abstract phenomenon. It breaks in predictable places, and the data makes those lines clear.
The personal price is the greatest. 70 percent of consumers say they would cancel, stop buying or leave negative reviews if they were charged differently for the same product.
There are also concerns about how recommendations are made. About 28% of consumers worry that AI is pushing sponsored products, and another 28% question whether the results are biased or irrelevant.
Control is another area of stress. Thirty-four percent of consumers are uncomfortable with AI ending purchases without consent, and 45% are uncomfortable with how their data is collected and used.
What worries you most about AI in online shopping?
AI is becoming a new layer of influence
AI is already shaping purchasing decisions, even if it hasn’t replaced human input. Omnisend found that 18% of US consumers prefer recommendations made with AI over friends or influencers.
That number is still small, but it could be indicative of a big change. For these consumers, AI selection is replacing search and social proof. AI-generated recommendations are becoming another area where brands clearly stand out or are overlooked.
If that share grows, it will cause major changes in the way marketers think about traditional channels like paid search, organic search and influencer marketing.
Clarity and control determine what happens next
The takeaway is not about how advanced the technology is. It’s about how clearly it’s explained and how much control consumers feel.
Consumers want to know what data is being used and why, understand why they are seeing certain recommendations and retain the ability to approve decisions before anything is purchased.
This is where most startups fall short. Even if the recommendations are accurate, a lack of transparency or control can undermine trust and reduce adoption.
See the complete picture of your search visibility.
Track, optimize, and win in Google search and AI from one platform.
Start a Free Trial
Start with
What does this mean for marketers?
Disclosure is now part of the experience. As concerns about sponsored endorsements increase, clearly labeling paid placements is necessary to maintain credibility.
A broad switch is straightforward. Consumers will share data and trust AI if the value is clear, but expect boundaries to be respected.
Success will depend on how much data is collected or how advanced the models are, and even more on whether the experience feels right, understandable and under the control of the user.
The full Omnisend report is available here. (No registration required).
Key takeaway: Generated by AI
- Consumers are willing to share behavioral data if it clearly improves recommendations
- From search to recommendation is already underway
- Trust breaks down in certain areas including pricing, transparency and regulation
- AI is emerging as a new layer of product discovery alongside search and social
- Clear explanations and user control are more important than technical expertise
