Most market research is built on a simple, broken transaction. A panelist spends an hour answering questions for roughly the price of a coffee, and a researcher gets back data where 60% of the responses were terminated halfway through [GreenBook]. The economics of survey sampling have, for years, incentivized speed and volume over quality, creating a fertile ground for fraud and apathy. Enlightn, a Montreal-based startup founded in 2025, is trying to rebuild that transaction from the ground up. Its bet is that by using AI to deeply profile real people first, and only then matching them to the right studies, you can fix the data,and maybe the incentives, too.
The match-first, activate-later wedge
Enlightn is an API-first platform, but it sells a process. Instead of the industry standard of blasting anonymous traffic at a survey screener to see who sticks, Enlightn works with panel providers to build rich, open-ended profiles of their members. It then uses semantic search to match these profiled individuals to specific study briefs. The platform only recontacts a panelist when there's a strong fit, aiming to turn the traditional high-attrition funnel on its head [Research Live, September 2025]. The founder, Adrien Vermeirsch, claims this approach can eliminate 60-70% of the termination rates that plague conventional sampling [GreenBook]. For sample buyers,the brands and insights teams,the promise is cleaner data. For sample suppliers,the panel companies,it's more completes per invite and less churn on their valuable asset: the panelist's attention.
A founder shaped by fraud
The company's thesis is deeply personal. Vermeirsch spent five years at survey platform Potloc, where his roles spanned research, product, and supply quality management [MRWeb, September 2025]. There, he ran research-on-research, studying how survey design and meager incentives corrupt data. He investigated fraud firsthand, tracking fabricated interviews and coordinated cheating in Telegram channels [Research Live, September 2025]. This front-line experience with the industry's dirty laundry convinced him the problem wasn't a lack of technology, but a misapplication of it. Enlightn positions its AI as "human-first," a direct contrast to the trend of synthetic respondents, arguing that understanding the real world requires asking real people,just the right ones [Enlightn blog, September 2025]. His work on these issues landed him a spot on the 2026 GreenBook Future List, a recognition for those working on data quality and trust in insights [GreenBook].
The incumbent inertia
The competitive landscape is dominated by volume players. Giants like Cint, Dynata, and Toluna operate massive, often commoditized panels where speed and scale are the primary currencies. Their models are optimized for the current equilibrium. Vermeirsch has publicly questioned whether paying respondents ~$2.50 an hour is sufficient for quality data, hinting at a belief that the entire incentive structure needs rethinking [LinkedIn]. Enlightn's challenge is to prove that its quality-centric approach can be scaled profitably within the tight margins of sample provisioning. It must convince panel companies to integrate its API and change their workflows, and ultimately, convince buyers to value,and pay for,higher-quality completes.
| Metric | Value |
|---|---|
| Survey Termination Rate (Industry Avg) | 65 % |
| Enlightn's Target Termination Rate | 20 % |
| Typical Panelist Hourly Rate | 2.5 USD |
What to watch in Montreal
As a very early-stage, founder-led venture, Enlightn's next steps are critical. The lack of disclosed funding or named customer partnerships suggests a bootstrap or stealth pilot phase. Success will hinge on a few clear signals in the next 12 months.
- First major supplier deal. Landing a recognizable panel provider as a client would validate the API integration and prove the unit economics work for the supply side.
- Published ROI metrics. The industry runs on case studies. Concrete data showing a 30-point reduction in terminations or a lift in data reliability will be essential for sales.
- Team build. Currently a solo founder operation, adding commercial and technical leadership will be necessary to move from a compelling service to a scalable platform.
On the back of an envelope, the math is stark but simple. If a typical $10,000 survey project loses 65% of its responses to terminations, that's $6,500 spent on wasted effort and fraud cleanup [GreenBook]. Enlightn's bet is that its matching engine can reclaim a significant portion of that waste as value,either as profit for the panel or as higher-quality data for the buyer. It's a bet on making the system less inefficient, one matched respondent at a time. To succeed, it doesn't need to invent a new market; it needs to beat Cint at its own game, not on volume, but on the actual cost of a reliable answer.
Sources
- [GreenBook] Enlightn | Data Collection from Canada - Greenbook.org | https://www.greenbook.org/company/Enlightn
- [Research Live, September 2025] Survey targeting company Enlightn launches | https://www.research-live.com/article/news/survey-targeting-company-enlightn-launches/id/5142723
- [MRWeb, September 2025] New Firm Enlightn Taps AI to Improve Respondent Quality | https://www.mrweb.com/news/new-firm-enlightn-taps-ai-to-improve-respondent-quality
- [Enlightn blog, September 2025] Enlightn is live: fixing survey targeting with human‑first AI | https://enlightn.io/
- [LinkedIn] Adrien Vermeirsch - Founder @ Enlightn | https://ca.linkedin.com/in/adrienvermeirsch/en