Awardee Retrospective: John Stockton

November 27, 2017
Bennett McIntosh
Livermore, Calif

The small signals that marketers, intelligence analysts, or environmental monitors need to find and act on can be covered by incredible amounts of noise. In 2012, John Stockton found himself at the head of two companies dedicated to finding those small, important signals — in very different ways.

Entanglement Technologies, which Stockton co-founded alongside fellow physicists Ari Tuchman and Hertz Fellow Tony Miller, developed from work they had done in graduate school regarding the quantum limits of detection. The company developed ultra-sensitive air quality sensors, which, five years later, played a critical role in tracking chemical safety in the aftermath of Hurricane Harvey — an experience Tony Miller wrote about for the Hertz Blog.

At the same time, Stockton and Tuchman were gathering a small team to launch a second company, Quantifind. Rather than sensitive hardware, Quantifind would focus on finding signals in all kinds of unstructured data. With rapid reductions in the cost of computing and information storage, and immense amounts of data readily available online, Stockton realized that clients from banks to marketers to the government were clamoring for ways to explain and understand the massive amounts of data they were collecting, and that a good machine learning platform could position itself to help clients in a wide variety of fields.

Finding traction in any particular field was a challenge: each kind of client has its own unique set of needs, and getting your foot in the door to prove you can address those needs can be difficult. Stockton credits his network of former Hertz Fellows, including Andy Huibers, among others, with helping him understand how to position Quantifind to break into new territory.

“Understanding our market, knowing when to pivot, was outside of our comfort zone in the beginning, as it is for most scientists turned founders,” Stockton recalls. “It’s great to have a network of personalities out there who have done risky things, and the Hertz Community was very supportive.”

While Quantifind initially marketed itself to scientists, legal firms, and government agencies, their first breakthrough was in the film industry, where fortunes are made and broken on the basis of opening weekend ticket sales. The challenge in the industry, then, is predicting and shaping customers’ behavior before a single ticket is sold at the box office.

To predict opening weekend sales, Stockton explains, there’s plenty of data out there. Social media, economic indicators ticket pre-orders — all may or may not hold secrets that predict the success or failure of a marketing campaign. But none of these potential signals individually tell the story that marketers can use to change their fate for the better. And even if there was a perfectly predicting algorithm, it’s fairly useless if it doesn’t explain why the product is going to underperform or overperform in particular areas. That’s where Quantifind comes in.

“Clients fundamentally don’t care about social media, or any other data source, in isolation,” Stockton explains, “they care about what impacts sales.”

Using what’s called “interpretable” machine learning — software which explains its reasoning rather than acting as a black box — Stockton and the rest of Quantifind learned how to tell the stories that marketers needed to hear. “We’ve since extended into verticals such as restaurants and financial services,” Stockton says. Among other things, Quantifind examines the impact that various sponsorship deals can have on sales — a notoriously difficult question to measure. On the flip side they have become adept at managing downside risk in crisis management, such as in fraud detection, and determining which data patterns do and do not justify sounding the alarm.

Wherever there’s data to interpret, Stockton hopes Quantifind can help people find the signal in the noise. “The technology goes beyond any one vertical,” he says. “We’re not a marketing company, we’re a ‘finding signals’ company.”

About the Peter Strauss Award

The Hertz Foundation Board of Directors established the Peter Strauss Award in 2010 through the endowment of Harold Newman, Hertz Foundation Director and financial advisor. Newman was a long-time associate of Peter Strauss, who died in 2010 at the age of 89. Both were supporters of the Hertz Foundation for many decades, serving on its Board and as financial advisors to the Foundation while senior associates at Neuberger Berman.