The TL;DR: In the past few weeks, conversations have arisen about the number of people researchers should interview. These conversations have been going on for decades, but they were recently spurred by the emergence of AI-native research platforms (like Genway AI ), which can interview hundreds or thousands of people per project. Questions have arisen: Isn’t that oversampling? Doesn’t quality matter more than quantity?

Our take? Researchers are OK at gathering data but fail at gathering diverse perspectives.

The diversity paradox in research

The business case for diversity is overwhelmingly clear. Research collated by LinkedIn suggests that gender and ethnically diverse organizations financially outperform their peers. Inclusive teams are more productive and make better decisions. In “The Diversity Bonus,” Scott Page demonstrates how cognitively diverse teams are more creative and successful at solving complex, high-density problems.

Diversity? Good.

Yet there's a fascinating disconnect: while (some of us) are (trying) to embrace diversity in our teams, we have not applied the same principle to our research. We are still primarily hearing from narrow slivers of 5-10 users because of a now antiquated (rather flimsy) study from 25 years ago.

Enter AI-Led (moderated) Research

This is where AI can be transformative by dramatically expanding our research aperture. Here's how:

1. Reduced Bias: AI can help research teams overcome a host of problematic biases.

2. Scale: AI can then scale research to less-biased samples, conducting hundreds or thousands of interviews simultaneously, reaching populations we typically miss due to time and resource constraints. For example: users in different time zones, users who prefer written to verbal communication, users who do not speak English and more.

3. Consistency: While human researchers might get tired after their fifth (and often last…) interview of the day, AI maintains the same level of curiosity and attention in the 500th interview.

How to get started

1. Audit your current research: Who are you hearing from? Where is your research suffering from bias or under-sampling?

2. Start small: Pick one underserved user segment and run a pilot AI interview project alongside your regular research.

3. Compare results: Look for patterns that emerge from AI-led interviews that might not show up in your traditional research.

4. Iterate and expand: Use what you learn to refine your AI interview approach before scaling up.

   
   

The Bottom Line

The most innovative companies are leveraging AI to hear more voices and base insights on strong foundations of diverse perspectives. By combining human expertise with AI-powered research tools, we can finally start building products that truly serve all our users, not just those most straightforward to reach.

You can get started today by going to https://www.genway.ai/try-it-free. Tap into previously unheard voices to build more customer-centric products.