As Facebook and Google continue to get richer off their users’ personal data, it’s fair to wonder whether they’re providing enough value in return. Is free access to Facebook’s social network an even trade when the average user in North America generates $112 per year? Is Google giving back its fair share when it generates $256 per year for each domestic user?
A new startup called Delphia believes it can provide a lot more value in return for that data, but not through ad revenue. Instead, Delphia wants to spread the wealth through the stock market, using personal information and direct surveys to make smarter investments and then paying dividends to its data contributors.
Andrew Peek, Delphia’s cofounder and CEO, claims that as the investment model attracts outside capital, an individual data contributor could make tens of thousands of dollars over the next 15 years, even without investing a penny of their own money.
To make such bold statements, Delphia is relying on some major assumptions about how well its model will work, how much capital it can raise, and how many users it can attract. But if everything goes to plan, it could be much more successful than other startups that solely want to pay users for their data’s advertising or market research value alone.
“Capital markets are 100 times the size of advertising, so if we were going to look for a place to co-locate our data and make it valuable, that is where we should look,” Peek says.
How Delphia works (in theory)
Delphia first launched out of Y Combinator last year as a way for websites and news organizations to offer AI-powered surveys, which would help readers work through complex decisions. The startup quickly realized there wasn’t much value in that approach and pivoted to investing. In researching the hedge fund business, Delphia discovered that some of the most successful ones were buying troves of data from sources like Foursquare and credit card providers to identify market trends.
“Most people are aware that data’s bought and sold in the context of advertising, but very few people are aware that there’s this practice amongst Wall Street to buy our data and use it to create an investment advantage for the 1%,” says Peek, who sold his previous startup to Shopify along with one of its main investors in 2013. A year later, he started working with Clifton van der Linden, a doctoral candidate at the University of Toronto, on building a business around forecasting real-world events, and they formed Delphia in 2017. Cameron Westland, Delphia’s CTO, joined as the company’s third cofounder last year.
Delphia’s approach is a bit different from Wall Street firms. While the company doesn’t rule out purchasing data through intermediaries, it’s also asking users to hand over data directly. Users will have the option to link their social media accounts, sign up for location tracking, and connect their banking history.
To supplement this behavioral data, Delphia will also send out one-minute surveys that either clarify users’ behavior (“Did you visit any of these stores while you were in Tribeca this weekend?”) or measure consumer sentiment (“What did you think of the Tesla Cybertruck, and how many of your friends might put a deposit on it?”). Peek says these surveys will give Delphia’s predictive models an advantage over Wall Street firms that only buy data in bulk.
“By combining the behavioral data with the questions we ask about you, we’re able to build out a much richer profile of who this individual is, which goes a much longer way when you’re trying to do machine learning,” he says.
Capital markets are 100 times the size of advertising.”
With that data in hand, Delphia’s machine learning models will try to identify trends in the market. Delphia employs seven data scientists along with a portfolio manager, who will then turn those trends into an investment strategy.
Delphia plans to charge a 2% management fee for any real money that investors put into the service. It will then distribute half of those funds to its data contributors. As Delphia’s return on investment grows, it plans to launch a hedge fund for accredited investors, at which point it will charge an additional “performance fee” that would also go to its data contributors.
In all cases, those who contribute more data—either by participating more frequently in surveys or connecting more of their online accounts—will get a greater share of those dividends.
While Delphia plans to distribute the payments on a weekly basis, it will require users to automatically reinvest those funds for at least 90 days before cashing out. The idea, Peek says, is to get people focused on the long view, and to see the benefits of compound interest.
“It is extremely important that they don’t look at this as income, and instead look at this as wealth generation,” he says. “This is what’s going to take care of their retirement or their kid’s tuition. If they’re treating it like income, they’re doing themselves a bit of a disservice.”
How much can Delphia make?
Delphia’s website currently provides a calculator that projects how much users might earn over time through the service.
If Delphia attracts 15 million users within five years, a user contributing as much data as possible—but no investment capital—could supposedly make $15,000 within 10 years. After 15 years, the total earnings could increase to $36,000. If that user also invests $100 per month in the service, the projected 15-year payout nearly doubles.
Very few people are aware that there’s this practice amongst Wall Street to buy our data.”
Bear in mind that these projections depend on some major ifs: Delphia assumes that it will have no problem attracting (and maintaining) millions of data contributors every year, and that those users will keep their earnings invested instead of cashing out. It expects to be managing upwards of $4 billion in assets within five years, even if only 1 million users have joined at that point. (Delphia says it’s planning to partner with publishers, talent agencies, and traditional investment advisors to promote the service next year, and it will also have a referral strategy.)
The biggest assumption of all, though, is Delphia’s expected annual returns of between 7% and 14% (before fees). Because Delphia doesn’t yet have the data to enable its supposedly superior investments, it can’t possibly back up this claim yet.
“I want to decrease the focus on the results right now, because we are now in the process of gathering the data” that will be necessary to get the best results, Peek says.
Delphia is gating participants behind an application process, meaning that it won’t start collecting users’ data until it has at least 100,000 contributors. But until that happens, it’s in somewhat of a chicken-and-egg situation: Contributors don’t stand to gain anything without significant outside capital, but that investment won’t come until Delphia can prove that its model works, and its model won’t really work without a critical mass of data contributors.
That might explain why Delphia is tossing out splashy numbers around its earnings potential before the product has even launched. The promise of thousands of dollars could be a powerful lure, even if the trade-off is a record of our banking transactions, social media activity, location history, and more. As Facebook’s and Google’s continued success makes clear, we’re willing to give up the same data for a lot less.