He ran the 400m for Germany. Then he tried to find the right expert for a McKinsey deck, gave up, and built a company that finds any B2B professional on Earth in a day. He calls it NewtonX. There are 1.1 billion names in the graph.
Sascha Eder sells time. Specifically, he sells the time between a research question and a research answer, and he sells the time of the specific pediatric oncologist or supply-chain vice president who happens to know that answer. NewtonX, the company he co-founded in 2017 with Germain Chastel, is the intermediary in that trade. It runs out of 185 Madison Avenue in New York, employs about 150 people, has raised $49 million across rounds, and does roughly $40 million a year in revenue. Its clients include Fortune 100 companies, institutional investors, and the same management consultancies that once used to send Eder on quests for the right expert and get the wrong ones.
The product is a search engine sitting on top of what Eder calls the Knowledge Graph, a data set of about 1.1 billion professional identities. When a client needs someone specific - not a "healthcare executive" but a specific procurement lead who has actually negotiated a specific drug rebate - the system contacts, on Eder's numbers, up to eighty thousand candidates in a day. The ones who fit are verified and paid. The ones who don't fit go back into the graph. There is no pre-recruited panel. There is nothing sitting in a database waiting to be surveyed. Every study is custom-recruited, which is a boring phrase for what is actually the entire product.
To understand why any of this is a business, you have to understand what B2B market research looked like before someone got angry about it. B2C research is a solved commercial genre. Nielsen counted what you watched. Millward Brown asked what you bought. The panels were pre-recruited because consumers are numerous and interchangeable enough that pre-recruiting works. B2B is different. If you need the perspective of a nephrologist at a specific integrated delivery network, or the buyer of industrial adhesives at a specific tier-two automotive supplier, there is no panel with that person waiting. The industry's answer for a long time was to retrofit B2C methods for B2B - to survey people who kind of resembled the target and hope. Eder's contribution to the discourse is to say, out loud and to the trade press, that this doesn't work.
He is not the first person to notice. He is, arguably, the first person to build a piece of software that assumes the observation is true and derives everything from it. NewtonX's refusal to call itself a panel is not marketing. It is the operating principle.
We don't really like the word panel. We understand what it takes to reach a certain type of audience - and we value that time fairly.Sascha Eder · Greenbook interview
Eder is German. He grew up training as a competitive middle-distance runner, made the country's national track and field team, and, by his own telling, kept the discipline afterward even when the racing stopped. He was a financial analyst at Procter & Gamble in Europe. He picked up a Master's in Management from HEC Paris. In 2014 he moved to the United States for a second Master's, this one from MIT. Then he did what a lot of MIT M.S. Management graduates do, which is take an offer at McKinsey & Company. He worked out of Seattle and San Francisco on tech-strategy work, mostly around AI.
McKinsey is where he met Germain Chastel, in 2015. The two of them kept running into a specific irritation, which was that when a consulting deck needed a primary source - a real expert talking about a real market - the sourcing process was slow, expensive, and often produced someone adjacent to the actual right person. They tried the incumbents. The incumbents were built for B2C. They tried again. Same result. So in 2017 they left to build a different thing, and the thing they built was NewtonX.
Eder brought BCG time to the résumé too, along the way. He built a strategy practice around AI at both firms, which is relevant, because NewtonX's operating premise is that you can use machine learning to identify a human expert faster than you can find them in a list. That is a bet about search, not a bet about panels.
The founding story is not a garage story. It is a consulting-firm story, which is different. It is two consultants who kept getting handed the same bad tool, quit, and built the tool. Eder tells it that way himself on LinkedIn - the origin, he said in a March 2024 post, was not entrepreneurship-for-its-own-sake; it was frustration with the specific pain of B2B research at a specific major firm.
Eder refuses to describe NewtonX with the noun "panel." This is not branding. It is a claim that pre-recruited pools structurally cannot serve B2B, and that custom recruiting is the whole product.
He was a national-team runner. He still runs. In interviews he talks about NewtonX in the language of pace and cadence, which is unusual for a founder who otherwise sounds like a consultant.
He and Chastel left McKinsey to build the vendor McKinsey should have been able to hire. The vendor is now selling to McKinsey-shaped firms.
Lived across three European countries before moving to the US in 2014. The company's Madison Avenue office is a five-country arc, essentially.
NewtonX takes its name from the physicist - a nod, in the founder's telling, to standing on shoulders and to a knowledge-first orientation.
NewtonX pays experts. Eder is loud about this. In an industry where a lot of "insight" comes cheap because it is compulsory, he treats compensation as part of the product's credibility.
The stated goal, in Eder's own phrasing, is "the world's leading B2B research company." That is a large, unfashionable ambition. Market research is not the sector most founders reach for in 2026. Its incumbents are old, its margins are respectable but not spectacular, and its brand associations are dusty. Eder's bet is that the entire category is due for a re-platforming and that AI is the mechanism.
The re-platforming, as NewtonX describes it, has three pieces. First, the Knowledge Graph - a data layer that treats every professional as a queryable object with skills, tools, industries, and verifiable identity. Second, a custom-recruiting engine on top of the graph, which for each study finds the specific people who match. Third, a delivery layer - dashboards, syntheses, longitudinal panels stitched from repeat participants - that turns raw responses into insight the client can act on. Eder's team also invests heavily in verification, because the entire pitch of B2B insight is that the person answering the survey is actually who they say they are. Panels are famously bad at this. Custom recruiting is, in theory, better at it, because the recruit is initiated from a verified professional identity, not a self-reported profile.
He is competing against a stack of incumbents including large legacy research firms, the panels that supply them, and the DIY survey tools that clients occasionally try before giving up. He is competing against fatigue, too - buyers have been sold on "AI-driven research" many times, and mostly by people selling much less than that. NewtonX's counter-argument is a specific one about data. The Knowledge Graph either exists and works, or it doesn't. The clients are the evidence, and the client list includes several of the largest technology, healthcare, financial-services, and private-equity firms in the world.
The size of the prize, if he is right, is essentially the whole of B2B market research. That is a multi-billion-dollar industry currently split among firms most consumers have never heard of. It is not a category Silicon Valley writes essays about. It is a category that spends a lot of money quietly.