(This story is from our new Health Tech newsletter. If you’d like to sign up, just click here.)
Among what seems like a gazillion startups using AI to solve administrative problems for healthcare providers, how does a company stand out from the crowd?
It’s a question Morgan Cheatham, vice president at venture firm Bessemer Venture Partners, has been grappling with. Health systems are exhausted with point solutions, and it’s nearly impossible for startups to go toe to toe with the deep-pocketed giants.
Take the AI scribe market, where there are some 40 companies vying for business from health systems and doctors across the US. Most of these companies would lose in a sales battle against Microsoft’s clinical documentation company Nuance, with its massive sales team and plenty of resources, Cheatham said.
So Bessemer is thinking about how to help startups focus on what they’re good at — building the product — while finding an edge with an innovative business model or go-to-market strategy to “address some of the shortcomings of a lower-resourced organization,” Cheatham said.
One method is to partner with an incumbent that could help bring the AI company’s tech to other customers while helping to develop its tools further.
“The traditional direct sales model in healthcare of ‘I’m going to go sell to every hospital,’ just doesn’t happen on a venture time scale in the way that many investors hope and expect,” Cheatham said. “And so we’re asking our companies to find ways to partner with incumbents that already have the existing distribution, the data, the know-how, the customer relationships, and really bring the technology into the folds of those organizations.”
Cheatham pointed to the AI scribe startup Abridge, which raised $150 million earlier this year from investors that included Bessemer, as one example of a startup partnering with an incumbent — the EHR company Epic — to not only distribute its tech to providers but co-develop new tools.
Another challenge facing AI startups is that healthcare providers don’t have the budgets to sustain large six- or seven-figure contracts for software. In place of software, Bessemer is considering whether an AI-enabled services company might make a more interesting business model, Cheatham said.
By way of example, there are tons of software companies vowing to help radiologists better analyze medical images. A more interesting business might be a teleradiology company that reads X-rays or MRIs for other clinicians, but uses AI to help its radiologists be more efficient, driving more profits and sustainability, Cheatham said.