strategies for building in highly intermediated industries (like education and healthcare)
Hello friends :-) Good morning and happy 2021! I’m resurfacing after a long break from publishing and would first like to request your feedback. If you have a minute after glancing through this, please reply to this and tell me what will make this newsletter most interesting for you, anything I can do more or less of, and any topics you’d like me to cover this year. Thank you for letting me share your inbox, and without further ado, hope you enjoy some reflections on building in two of the hardest and most rewarding markets, education and healthcare.
Applying software to highly regulated and intermediated industries like education and healthcare is tricky.
When well-meaning technologists dive into these industries they slam into a few major challenges. Human systems prove harder to hack than technical ones.
️️️🥊 1st challenge: more than usual inertia around the status quo. There are often multiple stakeholders who have fought to build distinct moats or for whom the status quo simply is profitable, and it distorts economic, productivity, and outcomes-based incentives.
🥊 2nd challenge: the disconnect between users, stakeholders, and payers. Where normally, design principles teach us to find a deep stakeholder problem and solve it, here you can do that and still not find a payer for it (sometimes never).
🥊 3rd challenge: systems need to be high availability with high privacy requirements. This is expensive to build to and means that unlike other areas of SaaS, you can’t spin up a product in weeks, deploy, and iterate.
🥊 4th challenge: slow sales cycles also make it hard for innovators to build quickly + iterate. A second-order problem: the fact that it’s so hard to build a high-growth business makes these unattractive to venture funders (for whom velocity is a major value driver)
🥊 5th challenge: “legacy infrastructure debt.” Like technical debt, except here it’s the effect of difficult-to-change systems around which other systems (including human systems) have calcified.
🥊 6th challenge: the people with the greatest needs have the least ability to pay. A common high-growth startup hack is to hone in on the segments with the greatest need AND highest willingness to pay, but that doesn’t work here. Many times those with the greatest needs have no ability to pay, and this distorts both the load and costs incurred by the system.
🥊 7th challenge: sometimes no one wants to pay to solve the biggest problems. As a society, we don’t have the collective will to pay kindergarten teachers more or pay for healthcare for the poorest people, so these problems linger and drag at the bottom line.
Things that work.
Despite the challenges, scrappy entrepreneurs find their way around them. But leaning into the problems help — as always, if you don’t acknowledge it, you can’t address it. It is tough to extrapolate successful strategies across two such disparate sectors, so please pardon me in advance. But I don’t want to give the impression that I think these sectors are hopeless — far from it. I’ve both built and funded here.
💪🏽 Finding discrete use cases: starting with what works and expand outwards. Very narrow use cases for fast adoption can provide a wedge into the larger stack.
💪🏽 Starting with discrete market segments where there are innovation-friendly, fast adopters: ie. self-insured employers, charter schools; or simply going DTC.
💪🏽 Rather than starting with the problem, you sometimes need to start with the payer and work backward. With ClassTag, advertisers are more likely to pay to get in front of parents than teachers are to pay for CRM software. With Flume Health, employers are willing to pay for cheaper, but high-quality, high-outcome health care, enough to disintermediate insurance companies.
💪🏽 A LOT of ecosystem help: in my observation, effective players are able to (somehow) mobilize the weighty, slow-moving, ill-aligned ecosystem in their support.
💪🏽 Patient and sector-focused capital: in these sectors, more than most, there is a huge advantage to finding funders who can wait, who understand the pitfalls, and who can help steer around the landmines.
There are NO worthwhile businesses to build that I can think of — which are easy.
So yes, human systems are harder to hack than digital ones. We are not 🤖 There are so many worthwhile and world positive businesses being built here — I’ve seen this firsthand as we have funded quite a few of them. But let it not be said, that it is easy.
Back to building ;-)
PS: don’t forget to drop me a note: how are you, what’s working, what’s not?