A New Era for Human Understanding.
Every few weeks, a number lands in your inbox or your feed. A satisfaction score. A political poll. A share of shoppers who prefer one brand over another. It arrives clean and confident, as if it fell from the sky. It did not. Someone had to write the question, reach the people, catch the answers, and turn a pile of messy replies into a single figure. For nearly fifty years, a lot of that work has run through software built in Montreal, by a company most people outside the research trade have never heard of: Voxco.
Voxco was founded in 1976, the same stretch of years that gave us the first Apple and the earliest Microsoft. Its job then was the same as it is now: help research teams ask questions and collect answers at scale. What has changed is everything around it - paper gave way to the telephone, the telephone to the web, the web to the phone in your pocket, and now to AI. Voxco is one of the few survey companies that lived through all of it and is still standing.
Most people meet a survey through a single channel - a link in an email, a call at dinner, a clipboard at a mall. Voxco's pitch is that a serious research team needs all of those at once, feeding one dataset. Its platform is built around that idea: Voxco Online for web surveys, Voxco CATI for computer-assisted telephone interviewing, Voxco Mobile Offline for face-to-face fieldwork that still works when there's no signal, and Voxco IVR and the Voxco Dialer for automated and predictive calling. On top of collection sit panel management, sample sourcing, and analytics.
The unglamorous part is also the point. Anyone can spin up a web form. Far fewer can dial a landline in rural France, run a mall intercept in Mexico, and stitch both into the same clean file that an analyst can trust. That plumbing is where Voxco has spent decades, and it is why the company keeps showing up inside the research firms whose names you do read.
Voxco's customers are mostly the people who make research for a living. Market research agencies. Corporate insights and customer-experience teams. Universities running academic studies. Government and public-sector bodies that need defensible data. The company says roughly three-quarters of the top 50 global market research firms use its software, and that the platform runs more than 50 million surveys a year across over 40 countries.
That customer base tells you something about the product's character. These are buyers who care less about a slick demo and more about whether the tool will still be collecting clean data at 2 a.m. on the last night of a fieldwork window. Reliability, not novelty, is the currency.
The problem Voxco talks about is the one every research team feels: clients want answers faster, the questions get harder, and the tools keep multiplying. The company's own framing is blunt - "say goodbye to the race for relevance." The trap it describes is a real one. On one side sit the big, slow incumbents that take forever to change. On the other, a swarm of shiny AI point tools that each do one thing and none of them talk to each other. Voxco's argument is that a research team should not have to choose, and should not have to become a systems integrator just to run a study.
Plenty of companies make survey tools. Voxco's difference is range. It handles phone research at real scale - CATI and predictive dialing - at a time when many rivals have quietly dropped it, betting everything on the web. That matters for the studies where phone still reaches people online forms miss: older populations, certain regions, regulated sectors. Voxco keeps both the old channels and the new ones under one roof, and offers the software on-premise as well as in the cloud, which suits buyers with strict data rules.
Survey infrastructure is not a category that trends. But that is precisely why it holds: once a research firm builds its studies, panels, and call scripts on Voxco, switching is expensive and risky. Unfashionable software with switching costs tends to age well.
The catalog follows the arc of a study. Collection comes first - Online, CATI, Mobile Offline, IVR, Dialer. Then the supporting cast: Voxco Panel & Portal for building and managing respondent panels, and Voxco Audience for sourcing sample when a team needs to buy respondents. Finally, analysis: Voxco Analytics turns responses into crosstabs, statistical tests, one-click reports, and dashboards.
The newest and most interesting piece is Ascribe, the text-analytics engine brought together with Voxco in 2024. Open-ended answers - the "other, please specify" box, the free-text complaint, the one-line rave - are the richest and most annoying part of any survey, because a human used to have to read and code them all. Ascribe does that with AI, and the company says it has processed more than six billion open-ended responses. That is the kind of number that only makes sense once you realize how many surveys quietly run through this system.
Voxco sells software to businesses - subscription and licensed access to its platform, priced by module, seats, and survey volume, with cloud and on-premise options. Around the software sit the services that research buyers actually need: support, professional services, and sample and panel sourcing. It is a classic enterprise model, where the initial sale matters far less than the years of renewals that follow once a research operation is built on top of the tool.
The obvious story in 2025 was that an AI startup would eat the aging incumbent. What happened instead was a merger. In November 2025, Voxco combined with Discuss, an AI-powered qualitative research platform, in a deal backed by growth investor Terminus Capital Partners. The combined company - roughly 200 people across 16 countries, serving about 900 customers - operates under the Discuss brand, with former Voxco chief executive Daniel Graff-Radford as CEO.
The logic is a familiar pairing. Quantitative research is about how many; qualitative is about why. Voxco brought the scale and the survey machinery. Discuss brought AI-led interviews and conversation analysis. Ascribe sat in the middle, reading the words. Graff-Radford's own background is a small detail worth noting: before software, he worked in medical research at the Mayo Clinic. He has been trying to get better answers out of people for a long time.
Raymond Cyr starts Voxco to build survey and data-collection software.
Sumit Aneja acquires Voxco through Pivoton Capital as the founder steps back after 40+ years.
An enterprise-software investor takes ownership to fund product and growth plans.
Daniel Graff-Radford becomes CEO as AI text-analytics firm Ascribe is brought into the platform.
Voxco and Discuss unite quantitative and qualitative research under the Discuss brand.
Voxco sits in the enterprise survey and market-research software market, alongside names like Qualtrics, SurveyMonkey, Forsta, Sawtooth, and Alchemer. It is not the loudest brand in that room. It has been the dependable one - the layer research firms build on rather than the logo they show clients. Its expertise is decades deep in the parts of research that do not photograph well: dialers, sample, panels, offline sync, coding open-ends.
The bet embedded in the Discuss merger is that the next era of research is neither pure survey nor pure AI, but a stack that does both and hands the analyst one clean picture. Whether that bet pays off is the open question. What is clear is that a company founded in 1976 has now placed itself, again, in the middle of how the world asks and answers its questions.