Why consumer healthcare is hard and how AI will change it

Why consumer healthcare is hard and how AI will change it

A quick summary on some thoughts and conversations I've been having

Neil Devani

Neil Devani

The healthcare AI hype has extended into consumer, which is going to be great for consumers, but I fear most of these business models will struggle.

A strategy I’m seeing is the “AI Doctor” offered at low to no cost, which drives top-of-funnel, and then you try to plug that into something “real” that can generate margin, e.g., lab work, virtual visits, or Rx fulfillment.

Last week, HHS published a roadmap that cracks open the door for AI doctors. Everyone was already rushing through it, but the margin on all of these activities is low and eroding. It’s a similar story to what we’re seeing with voice agents, where immense competition is great for customers but value-destructive for the 98/100 startups that don’t make it. Consumer AI companies are raising more money on bigger promises, which will drive more aggressive competition when consumers are already much more fickle.

The first strategy, lab work, has three challenges. First, labs are generally low margin. Second, patients will eventually recognize recurring labs aren’t necessary or convenient and churn. Third, the actual labs and their OEMs are oligopolies that can easily flex and take margin from the value chain.

The virtual visit business, whether synchronous or asynchronous, has also proven tough as a margin expander. Providers are protected by regulated supply constraints, i.e., too few doctors and hospitals vs. actual and potential patients. Expanding supply with AI lowers cost, enabling more price competition, further pushing down startups’ margins, which are already compressed due to clinicians having leverage on wages. Recurring virtual visits also haven’t been successful, given patients churn if they don’t have a real issue or escalate to IRL if they have a real issue.

Even the recurring Rx model is seeing margin compression. First, margin on logistics or scale is Amazon’s opportunity, especially as they’re figuring out virtual visits. If Costco mail-in pharmacy ever gets their shit together, that’s another 50-100M patients served with little to no margin. The IP holders are going DTC, with brand names taking more share in DTC businesses, both taking margin off the plate of startups. Verticalizing into compounding with peptides seems like a panacea, and if they run the supplements playbook, maybe, but to me, that’s not real healthcare.

By now, you might fairly assume that I believe the only opportunities are in traditional healthcare. But I believe in the consumerization of healthcare! I see consumers spending more than ever. The next phase will be consumer discernment, as they get more sophisticated. AI will quickly become table stakes and erode margins. Companies that can run super-tight ships, eking out margins via operational excellence and/or creating another kind of consumer surplus, will be able to earn spreads on someone else’s IP or clinical talent/licensing. But most will fail at that, and it’s hard to imagine that differentiator turning into a durable advantage absent Amazon/Costco scale.

Great businesses create more value than they capture, often much more. Thus far, the value creation in consumer health has been too limited, and mostly convenience-oriented. Talking to a doctor is valuable, as is getting a diagnosis. Solving the patient’s problem is more valuable, i.e., a treatment of some kind. Marginal improvements in cost, speed, quality, etc., to solving the problem can be valuable, but often only marginally. Solving a problem that isn’t already solvable is and will remain the most valuable. Luckily, there are tons of those still to solve! Getting patients/consumers to do things they say they want to do but don’t actually do is a $1T problem that AI will soon be better at solving than humans.

In a future post, I’ll profile companies that have done well, what’s behind their success, and some of our portfolio companies in this segment.

Thanks for reading! Subscribe for free to receive new posts and support my work.