Sana Benefits, an Austin company that wants to disrupt the health care insurance industry with more efficient software, said it has raised $6.3 million in seed funding.
The company joins a raft of upstarts remaking a notoriously inefficient health care system. These startups include players like Oscar Health, Clover, and Bright Health — all of which have raised hundreds of millions to billions of dollars over the last few years. Most of them have been focused on individuals and not as much on the employer plans that cover half of Americans.
Sana distinguishes itself in that it’s one of the few new companies focused on employers, and specifically the market for small and medium sized businesses (a thousand or fewer employees). This small business market covers a third of all employees — for a total market north of $200 billion, according to Sana CEO Will Young. Collective Health, which has raised more than $400 million, also targets employers, but only large enterprise customers.
The company’s emergence is the latest example of how massive industries like health — which makes up a fifth of the U.S. economy — can still be disrupted by applying what by Silicon Valley standards is modern, off-the-shelf software. In Sana’s case, it has taken things like Amazon’s HIPAA-compliant cloud architecture, Rails, and React and replicated incumbent industry data systems into a single web-based application — one that automates tasks that most other companies are still using manual labor for.
“Health care in general is a train-wreck of an industry,” says Young. “It is in the Stone Age.” He estimates that wasteful spending costs the U.S. health care system roughly $1 trillion a year, much of it from superfluous administrative costs.
Because of its ability to automate, Sana can operate with lower costs of administration, allowing it to drop rates 30% lower than market averages, said Young.
The lower costs, says Young, has helped it land dozens of customers (including Door and Abodewell), totaling thousands of members. It’s earning millions of dollars in revenues, up from nothing last year, according to Young. It offers free premium services to customers through partnerships with Plushcare, Maven Clinic, Beam Dental, ClassPass, and Calm.
The funding comes from Gigafund and Trust Ventures, which just invested $3.6 million, bringing total seed funding to $6.3 million. Both of those funds have focused on investments that aim to transform industries, said Young.
Sana can lower costs by using automation to adjudicate claims in 97% of all cases, meaning no humans are needed for the vast majority of cases — up from 0% when the company launched.
The ease of automation derives from the company’s decision to build its software from scratch, pulling data from incumbent industry systems.
That contrasts with most of the rest of the industry, where insurers usually rely on bloated legacy applications that were typically coded in the 1990s and aren’t compatible with one another. Most insurers employ different pieces of software for things like underwriting, accounting, member enrollment, and claims processing. These systems use clunky FTP servers and legacy data formats, where data is often delayed. Thus a single claim can be represented inaccurately across these systems.
Sana, by contrast, has integrated the data so that anyone looking at it “sees a single source of truth,” says Nathan Hackley, cofounder and technical lead at Sana.
To wring efficiencies from its claims process, Sana employs only two people in its claims department: one, the head of claims, pays the claims, and an engineer works alongside that person to automate all similar future claims. So in effect, the head of claims is really paying the claims and all future claims that look like it. The company can persist with just two employees for a long time, says Young, whereas other companies with comparable business might be using six or seven people.
Sana takes this same efficient approach across the business — pairing operations staff with engineers who automate. In Sana’s underwriting practice, for example, there’s only one full-time person. That person is doing the equivalent of what other companies employ 12 people to do, says Young. “We are just pulling in employee information, and the system spits out the right answer … we’re not sending Excel spreadsheets back and forth.”
Young said building software from scratch took a bigger up-front investment — about a year and a half of tech work, and $1 million in capital for setup costs — with tens of thousands of lines of code to vertically integrate the health care insurance software stack. This involved regular meetings with an industry consultant to make sure Sana properly coded edge cases required by regulations.
So far, the company has eschewed using AI or machine learning, in part because of the black-box problem of most AI models. “You put data in, and it’s hard to know why the system provides its answer,” says Hackley. But it’s also because there is still a finite amount of data related to each claim, making it easier to write a rule-based system for automation, Hackley adds.
Finally, on the customer-facing side, Sana has partnered with data vendors to calibrate rates more efficiently than the traditional method of using actuarial tables. It works to steer people to the best hospitals based on data analysis, and using telemedicine in cases where it is more cost effective.