Gabb + Qualtrics

As a leader in safe tech for kids, Gabb wanted to ensure its product and marketing decisions were grounded in a deep understanding of parents' concerns and behaviors. They faced a common challenge with traditional research: it was slow, sometimes expensive, and at times struggled to provide a clear view of niche audiences. They needed a faster, more cost-effective way to get reliable insights to support their brand and product strategy, without sacrificing data quality.

To address this, Gabb partnered with Qualtrics on a research program using a synthetic panel. The program’s objective was to compare how well AI-powered synthetic data would align with human panel responses on parents' perspectives about children's tech use. The goal was to determine if this new approach could provide a trustworthy, scalable solution for their pre-research testing and audience understanding.

The program focused on key areas of parental concern, including device ownership, parental control triggers, and attitudes toward social connection and screen time. The survey was fielded simultaneously to a traditional human panel and used to model synthetic responses, enabling a direct, side-by-side comparison.

Key takeaways and a path forward

The program successfully proved that synthetic data can provide reliable, directional signals for a fraction of the time and cost. The results showed broad alignment between synthetic and human data on core measures, including most means, top-box scores, and overall distributions.

Key learnings included:

• Directional Accuracy: The rank order of what triggers a parent to tighten controls aligned well across both groups, providing a reliable basis for strategy.
• Cultural Radar: The synthetic data served as a "cultural radar," accurately mirroring public discourse on topics like the "anxious generation." This provides a more cautious, yet valuable, baseline for stress-testing messages and safety narratives.
• Optimizing the Research Approach: The program confirmed that a blended approach is optimal. Gabb can now use synthetic data to quickly narrow down ideas or triage concepts, then validate the most promising ones with human panels for high-stakes decisions.

Embracing nuance: How AI data complements human research 

While some predictable differences emerged in niche demographics and questions about subjective topics like confidence and uncertainty, these variances were explainable and offered new strategic advantages.

For example, the synthetic data's tendency to avoid "I don't know" responses provides an excellent tool for building hypotheses and aligning stakeholders on key themes. The program moved Gabb from simply dabbling in synthetic to becoming a synthetic explorer—seeing clear applications for early-stage testing and trend detection.