When AI Gets the Customer Wrong
AI can help get to ideas faster. But speed and convenience don’t guarantee accuracy. An answer that looks correct at a high level can still get the details wrong, steering targeting, positioning and creative off course.
A recent STRAT7 study illustrates that gap.
The headline looked right. The details told a different story.
STRAT7 compared results from two synthetic data providers with actual survey responses from ~ 3,000 UK adults regarding snack-buying habits.
Overall brand awareness results were close. Deeper analysis revealed weaknesses:
Market changes were missed. AI identified year-over-year brand awareness movement correctly only 19% of the time across 16 brands. While real awareness increased substantially for many brands, the synthetic results barely moved.
Answers contradicted one another. In one provider’s pricing exercise, fully synthetic respondents gave inconsistent price rankings 68% of the time. Adding real respondent data reduced the inconsistencies but did not eliminate them.
Audience differences were misrepresented. Responses aligned overall on whether the World Cup encouraged snack purchases. Among women ages 35–44, however, one provider estimated 14%, compared with 31% among real respondents.
Purchase drivers were misidentified. One provider ranked store availability as the leading influence on purchase, although it placed sixth in the real respondent data.
Source: STRAT7’s July 2026 summary of its synthetic data study.
For agencies, the risk is that an AI-generated assumption becomes the brief, the brief becomes the creative, and the client invests in a campaign built on a flawed understanding of its customers.
Relying on AI without validation can turn an early error into a costly strategic decision, targeting the wrong people, emphasizing the wrong benefit and overlooking the opportunity that real customer evidence would have revealed.
Test the assumptions that shape the work
Before AI-generated insights become the foundation of a brief, test the assumptions that matter most. Focused interviews, quantitative research and concept testing can establish whether a proposed direction holds up with the intended audience before making a substantial investment in creative development.
AI can expand the possibilities. Testing provides the evidence to decide which ones deserve investment.