Free Ebook
From Access to Trust
What twenty years of online sample teaches us about trustworthy research, by Kurt Knapton
Kurt Knapton, former CEO of e-Rewards and Research Now and a Rep Data board member, looks back on two decades of online sample evolution. From double opt-in panels and river sampling to marketplaces, fraud prevention, and AI-driven quality approaches, the guide traces how the industry's scarce resource shifted from respondent access to respondent quality.
Get the free ebook
Fill out the form and the download unlocks instantly.
What's Inside
Five chapters, from the panel era to AI
The story starts in 2005, when trustworthy research was built on recruitment and panel management, then follows each major shift that reshaped what quality means. Every chapter ends closer to today's question. Can we trust the respondents we find?
Chapter 01
How the panel era defined quality
What did quality mean when panels ruled?
- Validate once, trust repeatedly
- Double opt-in and email verification
- Identity-focused, not behavior-focused
- Long-term panelist relationships
Chapter 02
Access changes everything
What happened when sourcing left the panel?
- Marketplaces, exchanges, and routers
- Respondents matched directly to studies
- No single supplier sees full activity
- Access solved, quality strained
Chapter 03
Pre-survey validation grows up
How did validation become a discipline?
- Digital fingerprinting
- Device intelligence
- Behavioral validation
- Cross-network visibility
Chapter 04
What quality looks like today
Where do quality decisions happen now?
- Before the survey with fraud screening
- During the survey with behavior checks
- After the survey with human and AI review
- Evidence combined across all three stages
Chapter 05
AI and the next quality challenge
What does AI change about trust?
- AI-generated and AI-assisted responses
- Proving someone is human
- Human benchmark data gains value
- Trusted data as the AI baseline
Twenty years ago the industry asked whether it could find enough respondents. Today it asks whether it can trust the respondents it finds. The marketplace model solved the access problem of 2005 while creating the quality problem of 2026, and the answer is layered validation across every stage of the research process.