The robots handle the saliva with care. They group scores of tiny, rubber-capped test tubes into orderly racks and position them under plunging needles. An articulated arm moves some of them from one automated track to another, and all of them eventually get loaded into a small, box-shaped device whose sole job is to vigorously shake things. This separates patients’ DNA from their spit. Everything that this huddle of roughly 200 robotic components is doing is about extracting valuable genetic signals from raw physiological noise.
The automated employees at this mechanical lab in South San Francisco are the clicking, whirring heart of Counsyl, a startup designed to make lifesaving genetic sequencing as commonplace as cholesterol tests. That means finding more efficient, and more affordable, ways of hunting down the telltale markers in DNA that indicate a patient’s risk of passing along certain diseases to a child, or of developing specific cancers herself. But it also means transforming the impenetrable and customer-averse field of medical diagnostics into something it’s never been–an inviting online experience. Robots are key to Counsyl’s strategy. Just as important, though, and maybe more revolutionary, is its user experience. Innovation, in genetic testing, can be as simple as treating your patients like humans.
If Counsyl’s bright, uncluttered website is your first time visiting a medical-diagnostics lab online, you might be duped into thinking they’re all so friendly. Typically, labs use their web presence as a kind of back-lot freight entrance, where requests, approvals, and other administrative communications are funneled to and from doctors’ offices, hospitals, and insurance providers. Customers simply get in the way. Counsyl, however, looks like an e-commerce site, greeting customers with clear information about the two DNA screenings it currently offers: family prep, which identifies whether you’re a carrier for more than 100 disorders that could be passed along to a child, and inherited cancer, which looks for your own elevated risk of developing breast, ovarian, or prostate cancer. Both tests have to be prescribed by a physician, but the bulk of the interactions, from placing the initial order to receiving a clear explanation of the diagnostic results, are between the patient and Counsyl. Whereas genetic-testing company 23andMe was founded with a mission to inform customers directly that they might have a predisposition for Parkinson’s, say, or Alzheimer’s, Counsyl focuses only on what it calls “clinically actionable data”–information that warrants direct treatment now, which could range from an early mammogram to checking an embryo for specific conditions before in vitro implantation. So far, the company’s fully accredited laboratory has tested roughly 250,000 people. The company currently works with more than 15,000 medical professionals and a network of insurance providers that includes some 150 million subscribers.
That Counsyl’s tests are presented as actual products–which you can choose to buy (with your doctor’s approval) rather than be instructed to undergo, is a radical departure for diagnostics. Even more radical is Counsyl’s online cost calculator. Each of the screenings costs about $1,000 (roughly 70% less than what competing companies charge). But enter basic info about yourself and your health insurance, and algorithms churn through more than 1 million data points, collected from the quarter-million insurance claims the company has already processed. Within moments, you’ll learn exactly what the service will cost you out of pocket–typically between $150 and $300–and how much your insurance will pay. “We wanted to bring price transparency to an industry that wasn’t known for it,” says CEO and cofounder Ramji Srinivasan, 32. This cost calculator is unique to Counsyl and central to its goal to become the first diagnostic company that patients actually want to interact with. While others perform similar tests, says Srinivasan, “nobody wakes up in the morning saying, ‘I love LabCorp. I love Quest.’ ”
In a conference room a couple of halls down from the lab, in an unassuming office-park building, product designer Laura Martini walks me through Counsyl’s user interface. There are no walls of impenetrable text and a minimum of lawyerly fine print. The tests are explained with slick animation and video sequences. Martini, who previously worked in product design at Apple, demonstrates how requesting approval from your doctor amounts to entering his or her contact info. Counsyl does the administrative grunt work, such as faxing paperwork and making follow-up calls. “One thing I learned at Apple is that what makes a good product is taking these really complicated systems and making them effortless and simple,” she says.
That applies not just to the online ordering experience but also to the oddly attractive saliva kit that users get in the mail. The matte-finish boxes have the same luxe feel as the packaging used for iPhones and MacBooks. Intuitive directions appear on the inside of the lid. It feels like a retail product. The prospect of filling a huge tube with your drool feels oddly aspirational. “When you think of some of the best companies, like Amazon or Google, once you have that user experience, it’s hard to go back to anything else,” says Aydin Senkut, whose firm, Felicis Ventures, offered several rounds of funding after he and his wife had a positive experience with the family-prep screening in 2011. “We have confidence that it’s a multibillion-dollar company.”
“When we started, we thought we were a software company,” Srinivasan says. In 2009, he and cofounders Eric Evans and Rishi Kacker–all Stanford colleagues–planned on being data-sifting middlemen, interpreting the results of highly detailed genetic testing but outsourcing the actual analysis to existing labs. It was a decidedly Silicon Valley plan, to create value through organization, without having to design or build anything physical. And it was a dead end.
The problem was the lab results, which were coming back effectively useless. “The tests that we were running required high dexterity,” says Srinivasan, over lunch in the company’s conference room. He was used to lab work, having run genetic sequences with Evans at the Stanford Genome Technology Center, and understood why Counsyl wasn’t getting the data it needed. “If you held certain samples with your fingers, you would melt the enzymes. It wasn’t something that we could outsource,” he says, explaining that the lab technology “was something that we needed to build ourselves.”
For a company with such a vibrant public-facing presence, Counsyl’s lab, which cost the company between $10 million and $15 million, is strangely lifeless. There are no white-coated technicians or geneticists peering at beakers and computer monitors. There are only robots, a cluster of fully autonomous machines sliding and shaking through their paces, small enough to fit inside most janitorial supply closets. They’re true robots, too, not simply looping through some preset, assembly-line sequence but adapting to input, adjusting their actions based on results, with no interference from humans. “They’re robots doing experiments, and then informing robots who are doing other experiments,” says Kyle Lapham, the head of automation. “We’re constantly upgrading the system. We’re years ahead of what anyone else is doing.”
It would take 100 humans to do the work that these robots do–using as much as five times the square footage. They’re what enable Counsyl to charge 70% less than its competitors for genetic testing. “Dropping the price takes it from a luxury good to something that can be routine and available more broadly,” says cofounder and chief scientist Eric Evans. A published PhD geneticist, Evans sees Counsyl as actively saving lives, while many academics continue to pour resources into sequencing whole genomes. “I just find it infuriating that that is their focus,” he says. “Why would you want to tell somebody about 22,000 different genes when what you really need to do is tell them about the 100 or 200 genes that they can do something about right now?”
That sense of urgency is built into everything Counsyl offers, especially, as you might expect, counseling. Normally, setting up an appointment with a genetic specialist to explore what a set of results mean, and what medical options are available, can take days or months, assuming those resources are even available in a given state. With Counsyl, the wait can be half an hour. After logging in and reading completed lab results (which are also sent to the prescribing physician), a patient looking for immediate feedback waits on average only 30 minutes to speak with one of the company’s two dozen genetic counselors nationwide, who have completed some 10,000 sessions in the last year alone.
This, according to Counsyl, marks the most significant and meaningful departure from how diagnostics are typically handled. With other labs, patients might get heartbreaking results by mail, or be told something cryptic over the phone, if any direct contact is made at all. According to Srinivasan, Counsyl’s own analysis has found that with other labs, some 7% of “abnormal” diagnostic results–where further action might be necessary–may never reach the patient. Since launching its robotic lab, none of the startup’s more than 250,000 genetic samples have been mislabeled or misplaced, and every analysis is backed up by digital confirmation of exactly when the results were read.
Counsyl’s biggest challenge now is merely to grow. The startup, which has more than $100 million in funding from investors, including a recent infusion of $28 million from Goldman Sachs, only unveiled its inherited cancer screening in May, and the response has been strong. But Evans wants more. In the long run, he’s hoping for “a throughput on the order of 10 million samples a year,” he says. At that rate, Counsyl could conceivably identify almost everyone in the U.S. with inherited cancer-carrier genes, estimated at 1 million. It could also mean screening the parents of all of the 4 million children born every year and finding ways to treat or avoid countless serious and life-threatening conditions. “It’s great for us. We’d love to be the technology company that does that,” says Evans. “But for medicine, it’s even bigger.”