Marketing to the Big Data Inside Us

Companies market to you according to your shopping habits, your age, your salary, and your social-media activities. In the future, they may be able to advertise to you on the basis of your DNA.

Do you carry the genetic variants associated with lactose intolerance? Here, Lactaid has a coupon for you. The genes for male-pattern baldness? That’s accelerated by stress, so maybe you should come in for a discounted massage Jimmy’s Spa & Bath.

A Minneapolis-based startup called Miinome plans to build what it calls the first “member-controlled human genetic marketplace.” The company, which has just three full-time employees and is still hunting for financing, is notable mostly for its bold idea: to sell DNA information to marketers.

Digital information about people is exploding, and DNA data is no exception. This year, as many as 250,000 people could get their genomes entirely or partly sequenced, according to some estimates. That figure, while still modest, has been climbing exponentially and will soon reach into the millions.

Still missing, though, are widely accepted ways to store, share, and mine the data for medical and commercial applications. Miinome’s idea is a membership club that electronically stores the DNA data of people who have had their genomes sequenced, who then get to decide what information is made available and for what purpose. “[It] could be anything from donating their information for philanthropic purposes, such as an American Diabetes Association study, to savings on certain products,” says Paul Saarinen, the advertising executive who is CEO of Miinome.

Widespread gene-based marketing isn’t feasible yet, but it could be in the future. One reason is that the cost of decoding just the most important parts of a person’s DNA—only those stretches that code for proteins, known as the “exome”—is only about $700, and that’s expected to decline. Miinome may eventually offer to sequence the genomes of people for free, so long as they participate in its database.

That could take human genetics toward an ad-supported model, just like social networks and search engines. James Ostheimer, the data scientist who is Miinome’s cofounder, says consumer Internet companies like Amazon and Twitter have developed the key technology that’s needed—big server farms, software to handle massive amounts of personal data, and algorithms to mine it.

One problem has been that while DNA sequencing is cheap, analyzing the data isn’t. Some of the work is still done manually, by PhD scientists who sift through tens of thousands of DNA letters looking for those that might be the cause of a disease. Dietrich Stephan, CEO of SVBio, a company in Foster City, California, that has developed software to analyze genomes automatically, says the volume of data produced by modern DNA-sequencing instruments has overwhelmed hospitals and diagnostic labs.

Furthermore, although medical researchers typically want to link genetic differences to specific diseases, common diseases like diabetes are influenced by many factors, including the environment. As a result, Stephan says, making meaningful health predictions from someone’s DNA will require “rigging up all data pipes,” including medical records, the pollutants and infectious diseases a person has been exposed to, maybe even grocery receipts. “Until we get that data all in one place, it’s going to be difficult,” he says.

Sean George, a senior executive at InVitae, a San Francisco company now offering a genetic test for 270 genetic conditions to a handful of hospitals, says too few people’s genomes have been sequenced—and shared—to power the kind of analysis researchers want to do. “At a low number, that data is not super useful,” he says. “We do believe there is a tipping point [coming].”

The entrepreneurs behind Miinome think consumers will start to share their DNA data when there are applications beyond medicine. “In general, we’d like our members to hook up as many feeds as they want to,” says Ostheimer. He thinks if people link their DNA to their Twitter or Facebook feeds it could be possible to “look for new associations”—say, between certain genes and a taste for spicy foods.  

Scientifically, that sounds like a stretch. But commercially it might not be. “That’s an interesting marketing opportunity [for] people selling groceries or trying to get you to eat in their restaurant,” says Ostheimer.