Best vs Most Integrated Product, Heidi’s Achilles Heel
Last week I summarised the Heidi Interview on LinkedIn:
1. Heidi doesn't try to do documentation for diagnostic purposes, which is already a very saturated and regulatory-heavy space. Instead, they make a complete transcript of what happened and use that to improve budget planning on the government/health fund side and decrease reimbursement-process costs (especially in the US).
2. Running your own model on your own hardware is 100x cheaper and more consistent quality-wise for task-specific uses cases (especially if you have millions of data points to work with). This cost advantage is why Heidi can easily offer a free version, which is why they were able to grow bottoms up without a sales team (though they still have those).
3. Product experience matter big time. The models might get commoditised with time, but thinking through and fixing the long tail of edge cases is what delights the customer and what creates an experience moat. (Just ask Apple. Also, this is what many academic projects completely miss.) This is the reason why clinicians prefer to use Heidi from the Heidi app/website and not from the EHR-integrated version.
4. You can totally hire your CMO from BigTech when you collaborate :)
5. BigTech (Microsoft) is not a threat to Heidi, because they have bigger fish to fry. I disagree: Healthcare is a big fish and Microsoft's second largest acquisition was of Nuance for what we now know as Dragon Copilot. They are very much onto it. I do agree that Microsoft generally sucks at product, but a platform play and a bundle play is very much their sweet spot. There is a reason why Teams won over Slack.
If you worked in tech, especially consumer tech, you know how important a great product experience is for your success on the market. I compared this understanding to the medical space in Prof Valmed Wins DMEA Award, Certify It And They Will Come, and concluded that product experience continues to matter here too. Heidi’s growth seems to be confirming that same rule, to their own surprise. The company serves 3 million patient visits across multiple markets, to a big extent driven by individual clinicians just loving and recommending the product. Having your own brand evangelists doing youtube videos is a milestone in its own right!
The counter argument is that long-term and in the context of enterprise sales - which is what selling to hospitals is - once a product meets a certain good-enough experience threshold, other features become more important. One such feature is security; another is the level of workflow integration. And this is where the Microsoft point above kicks in.
It all comes down to perspectives. The workflow is a multi-step, multi-person process. Saving an hour out of a clinician’s time is really impactful for that clinician, which is why they are happy to adopt Heidi. But seen from the hospital lens, this one hour is a small slice of a big pie, that may even cost more than it brings, as speeding up one step without speeding up the rest of the system creates bigger bottlenecks and ultimately breaks the (work)flow.
Heidi and other companies know that, which is why they are racing to integrate as vertically as they can, the question is do they stand a chance? Slack is a telling story here. For a while they stole the hearts of individual developers, only to ultimately lose to a worse product (Teams) which was part of an all-in-one bundle, which was one click away from integration thanks to Microsoft’s ultra-wide distribution channel.
In the startup world the advice is to start with a niche, laser-focused approach and avoid the “all-in-one” product, where small companies can’t compete anyway. This is what Heidi has done, and done it right. Now that they have accumulated some speed, they are going all-in, racing to build their own bundle enough to be competitive with the big guys, disrupting the giants before they wake up. But this is a game that Microsoft knows how to win and one where they are not really snoozing on either. The game is on!