A bicycle race, Scott McKinley likes to point out, does not lie. You ride 120 or 130 miles, and the first rider across the line is the winner. There is no spin, no adjustment, no story to sell afterward. "Cycling is the ultimate test of truth," he says. It is a strange sentence coming from a man who now runs a software company in Oakland - until you learn that the whole company is built on the same idea.
McKinley was born in Seattle in 1968 and grew up fast on two wheels. By 1986 he was finishing eighth at the World Junior Road Race Championship. Two years later he captained the United States Olympic Road Cycling Team at the Seoul Summer Olympics. Through the early 1990s he raced for the biggest names in American cycling - 7-Eleven, Motorola, Coors Light, Saturn - collecting stage wins in Europe and national-championship podiums at home.
Then he quit. Not because he lost, but because winning stopped meaning anything.
Doping was spreading through the peloton, and the purest test of truth McKinley knew was quietly being rigged. He could keep racing against people who cheated, or he could get off the bike. He got off the bike. It is worth holding onto that decision, because he made a version of it again, later, in an industry that could not be further from a mountain pass in the Alps.
From TV station websites to Nielsen
The second act started small. In the late 1990s McKinley managed websites for Cox-owned television stations - unglamorous work at the dawn of the commercial internet. But it put him at the intersection of media and data just as that intersection was about to become one of the most valuable pieces of real estate in the economy.
Over the next two decades he co-founded and ran a string of measurement and analytics companies, then landed at Nielsen as an Executive Vice President leading global product innovation - person-level data, identity management, location services, and the acquisitions that stitched them together. After Nielsen he ran an identity data firm called IDify. By any measure, he had become an insider in the machinery that decides which ad reaches which person.
And the longer he sat on the inside, the less he trusted what he saw.
The snake-oil problem
Digital advertising had grown into a business measured in hundreds of billions of dollars, and much of it ran on data about who people supposedly are - their age, their income, their intent to buy a car next quarter. The uncomfortable secret McKinley kept bumping into was that a great deal of that data was wrong, and almost nobody was checking. Vendors sold "audiences" with confident labels attached; buyers had no independent way to know whether the labels were accurate or invented.
He put it more bluntly elsewhere: he was "so tired of the snake oil salesmen, the obfuscation and BS" that he nearly left advertising for good. He is especially withering about probabilistic modeling - the practice of guessing attributes at scale - which he calls "basically a euphemism for pure guessing, with an incentive to maximize scale at any cost." The industry, in his telling, got addicted to scale and let precision rot.
This is the moment where the cyclist's instinct returned. Faced with a game he believed was being cheated, McKinley could walk away again - or he could stay and try to build a scoreboard. In 2019, he stayed.
What Truthset actually does
Truthset does something deceptively simple: it grades data. It does not sell data. It is not a data broker. Instead it assembles a "Data Collective" of more than 20 data providers, pools their records, and uses machine learning to produce an accuracy likelihood score - a TruthScore - for individual consumer records. A buyer can then see, before spending a dollar, how likely a given data set is to be right.
Think of it as an independent lab certifying a sample, or a referee standing at the finish line. The value is not in owning the data; it is in being the trusted, disinterested party willing to say this record is probably true and that one probably is not.
Crucially, McKinley frames it as a rising tide rather than a fight. "This is not a zero sum game, where there's winners and losers," he says. Better accuracy helps data providers prove their worth, helps publishers and brands spend smarter, and - not incidentally - treats the consumers behind the records as people rather than proxies. That last phrase, "people, not proxies," has become something of a rallying cry for the company.
The long ride, in dates
Why the cycling metaphor sticks
It would be easy to treat the Olympic backstory as a colorful footnote. It is more useful as a key. McKinley keeps returning to cycling because the sport gave him a clean definition of integrity: the result is unarguable, and cheating is a choice that corrupts the whole thing. When he says advertising has a data-quality problem, he is describing the same corruption he skated away from in the 1990s - just wearing a suit instead of a jersey.
His skepticism extends to fashionable practices like ID bridging, where a single person's identifiers get stitched across devices in ways that inflate reach and blur accuracy. In a world losing third-party cookies and hunting for new ways to recognize people, McKinley's pitch is almost old-fashioned: measure what's true, admit what you don't know, and stop rewarding scale for its own sake.
In October 2024, Truthset closed a $5 million Series A, pushing its total raised to roughly $9.75 million - modest by the standards of a hype-driven industry, which is arguably the point. In mid-2025 he was still on the road, appearing at Cannes Lions to talk through phantom profiles and flawed identity graphs with media analyst Evan Shapiro. The message hasn't drifted. The finish line is still the finish line.
There's a certain symmetry to it. A young athlete once refused to keep pedaling in a race he no longer believed was honest. Decades later, handed a much larger and messier race, he chose the opposite move - not to leave, but to build the referee the sport never had.