Your AI ads may feel too personal, Snapchat’s survey suggests

The April survey separates enthusiasm for AI help from comfort with brand personalization.

Snapchat's company-supplied AI Sponsored Snaps example showing an Experian chat across four phone screens

Snapchat’s latest research roundup carries a warning for AI advertisers: 70% of the consumers surveyed considered it intrusive when brands used personal data to create hyper-personalized AI ads. The finding comes from people who already use AI.

Snapchat highlighted the findings on October 22. Ipsos conducted the research for Snap and WPP, asking 4,261 consumers aged 18 to 44 in the US, UK, France, Saudi Arabia and India about their use of the technology.

The full report dates the online fieldwork to April. Respondents used social media daily and had used AI in the previous month; the results describe that selected group, rather than all consumers.

Comfort changes with the job AI is asked to do

There was considerable enthusiasm for everyday assistance. In the consumer sample, 48% said they used AI daily, while 54% used it to help express thoughts or feelings to friends when they couldn’t find the words themselves.

Survey chart showing 54% use AI to find words, 53% for creative communication and 62% to keep conversations going
Snap's chart describes communication uses among the study's selected AI-user sample, not the general population. Image: Snapchat, Ipsos and WPP

Delegating a purchase drew less interest. Some 62% said they were likely to use AI to curate media or events, compared with 37% for managing purchases. Those are stated preferences, rather than measurements of completed transactions.

Survey chart comparing 62% willingness to use AI for media or event curation with 37% for purchase management
Respondents expressed more interest in AI curation than purchase management. The survey measured stated willingness, not observed transactions. Image: Snapchat, Ipsos and WPP

Customer service had another clear condition: 68% considered it a deal-breaker if a brand’s AI chatbot could not connect them immediately to a person. That gives brands a specific point to examine when designing a chatbot handoff.

Test the data request alongside the recommendation

For an advertiser building a shopping assistant, the findings suggest testing how much information people will share at each step. A visitor who welcomes help choosing a product may still object to an ad built from personal history.

For example, a gift finder could ask for a budget and the recipient’s interests, explaining how those answers shape its suggestions. Teams could compare variants requesting different details and track completion alongside product selection.

That is a proposed test, not an outcome measured by Ipsos. The survey does not establish which personalization choices increase sales or how respondents actually behave when shown an ad.

Snap sells AI-enabled advertising and uses the research to recommend its own formats. Its advice to offer clear personalization choices can be evaluated separately from those product recommendations.

Source: Snapchat

Want SEW higher in your Google results?Add as a preferred source

More in Analytics

View more
AI Search

Your ChatGPT traffic may be hiding in Direct

Missing source information on some ChatGPT links is complicating referral reports, while rank trackers are dealing with a separate problem in Google results.

Start the conversation by posting the first comment

Join the conversation

Posting publicly · your email is never shown