Completely new healthcare markets and what to build for them
Get Out-Of-Pocket in your email
How to Build a Healthcare Call Center
That new market smell
I’m going to do a series of emerging areas in healthcare over the next few weeks (or months, idk stop rushing me). Today’s post is going to start with companies in brand new markets that were virtually non-existent 5-10 years ago. In subsequent posts, I’ll go through new business models and new users.
I like to invest in B2B or B2B2C companies where the customer segment is in a newly created market that’s small but growing. This is for a few reasons:
- They’re usually underserved because the market is so new and small that incumbents have decided it’s not worthwhile to build for the segment.
- It means lots of new customers are about to make their first purchasing decision for a problem they encounter. A first time buy is much easier for a new vendor to win vs. a rip and replace situation.
- You can grow alongside the customer as they grow themselves - which creates lots of upsell opportunities.
- There are usually strong word of mouth effects from these new customer segments because there aren’t many of them, and usually they’re navigating uncharted waters, which means communicating/asking a lot of questions from peers about vendors they use.
- Offering services can be an excellent wedge into product onboarding since they’re usually looking for all the help they can get vs. large companies where you might need to navigate internal politics of people building your solution internally.
- You can laugh at your haters when you turn out to be right.
There are also downsides to picking this strategy.
- You have to guess correctly that the number of customers in the segment is going to increase and the new, small market is going to grow into a large one.
- The contract sizes are much smaller in the beginning and generally less sticky for customers.
- When a downturn happens, you can be overly exposed to a customer segment that evaporates or needs to cut down on their operating costs
- Competition becomes fierce once the segment grows and you’ll have to demonstrate you’re not just a feature but a necessary standalone application. Incumbents with more distribution, other products to bundle, and aggressive sales forces usually start getting involved.
- You look like a fucking weirdo interested in fringe shit no one has heard of or cares about. You become a social outcast.
But in general, I still think it’s a good rule of thumb to sell something to customers in small, emerging markets. You can look at analogs in other industries to see where this strategy has been successful. Affirm offered new payment options initially targeted at e-commerce which grew rapidly, Plaid offered authorization rails for the fledgling fintech industry, and Unity built a gaming engine that got a foothold with indie developers and mobile game developers before expanding to large studios.
Here are some small but growing customer segments in very new markets I’m watching. And by watching, I mean “read about like twice”. I try to list out some of the unique needs of these segments that vendors are stepping in to solve, but would love to hear what I’m missing from people that actually work with these customers.
Because of the nature of this post, lots of companies I’m invested in or advise are mentioned. They are denoted by the ** next to them.
Remote first companies
A lot of the companies that got started during COVID had to be remote-only by necessity and many are staying that way. Plus, even companies that were in-person have become much more flexible about remote workforces. What they gain in an expanded pool of talent untethered by geography, they lose in not smelling someone heating up leftover salmon in the common area.
Employer health benefits are not really designed for remote workforces. Provider networks and insurance choices are meant to service employees clustered in specific states, and typically, vendors and employees’ enrollment programs focus on the office as a main hub. All that work of creating a narrow network of high quality providers near your headquarters, only for everyone to move away. Those benefits consultants are punching the air right now.
With remote-first companies, not only do these need to be re-thought, but we might even have the opportunity to offer new benefits.
- Setting up wallets or payments for employees in different states and countries to get their own health insurance in their locale. Lunchbox and SafetyWing are creating products for this.
- Offering care coordination options for employees to find high quality providers wherever they’re living for a given issue.
- Designing virtual enrollment and awareness campaigns for employees that aren’t in the office. Companies like Fertu are tackling this.
- Sending flu shots, blood draws, and other screening tools to patients’ homes, especially for ones at risk. Many friends that work remotely with their kids at home now can’t leave in the middle of the day to get labs, etc. Axle** is making it easier to deploy clinical staff to a patient’s home to do this.
- Creating virtual wellness programs with leaderboards, etc. or making it easy to enroll in a virtual program of the employee’s choice. I want my coworkers to know I have a workout video on even though I’m kinda scrolling Twitter between sets and not pushing myself as hard as I should.
- Making it easy to set up domestic medical tourism. With every state deciding their own coverage rules (with abortion being one of the highest profile examples of this), employers will likely want to make it easier to get their remote employees care wherever they want to get it. For example, Amazon offers this for abortion, gene therapies, substance use disorder, etc.
Unfortunately the employer-based healthcare system we have means that when employment structures go through a massive overhaul, the health benefits need to as well.
Healthcare AI startups
Lots of new startups are building companies where their AI/ML models are a key differentiator. If you believe AI is going to be a meaningful part of healthcare going forward, it’s worth figuring out what these companies struggle with as they build or what they had to build in-house that was outside of their core competencies. If you don’t, you should continue enjoying your cave paintings and the advent of fire.
What are some things the AI companies themselves would need to buy? Some examples:
- Data pre-processing and standardization is a huge part of the data janitorial work that AI startups have to do and many end up building their own custom ontologies. Tuva** is building open-source software to standardize some of those ontologies.
- Annotating training data and making it queryable for companies that want to find the right data to train their models. Gradient Health**, Centaur Labs, Medcase have some combination of annotation tools for labeling, a workforce to actually do the labeling, and a way to query for the specific type of patient/disease pathology you’re looking for across the aggregated dataset.
- Making it easy for companies with data to make it commercially viable through de-identification, cleaning, privacy preserving techniques like federated learning, etc. so that it can be bought or trained on. Rhino Health seems to be playing in this space.
- A data marketplace that makes it easy for companies that want a full picture of patients across different data types like EHR, imaging, pathology, etc. to request that data for training. You need a rich clinical context about a given patient when training a model, and it’s rare that you can rely on a singular dataset to give you what you need. Individual hospitals seem to be doing this (e.g. Mayo Clinic has their Accelerate program) and in countries with centralized health records, this is common (e.g. Israel has invested a considerable amount into this). Companies like Dandelion Health are working on this, and maybe some of the data packagers like Komodo end up getting into this.
- Cloud costs for these startups are going to be high, so figuring out ways to optimize spend is going to be key. The main cloud service providers have this plus companies like Cast.ai work on this. Is there a healthcare-specific version of this/does it need to exist?
- Dealing with AI regulatory approvals and reimbursement seems to still be fuzzy for software-as-a-medical device companies. Figuring out how to go down the De Novo approval pathway, keeping compliant during software updates, and figuring out how to navigate things like the New-Technology Add-On Payments (NTAP) or set up value-based care contracts could be useful.
Healthcare Crypto Companies and DAOs
I’m bullish on the potential of crypto companies in healthcare. Unrelatedly, I’ve discovered whippets.
But seriously, I think crypto has the potential to solve some of the problems that particularly plague healthcare, including agency over data, coordinating groups to an end goal without relying on a middleman that charges insane take rates, and ownership in projects which we contribute to.
But the reality is that this space is still so early. Decentralized Autonomous Organizations (DAOs) are of particular interest to me, and I talked about them in a previous article about peer-review. At a high level, these are essentially organizations without any leaders that try to accomplish goals through proposals, voting, budgeting, and more. Some of these use features of crypto and blockchain technology, and some don't.
I’ve been playing around with a lot of the existing DAO tools since then and…they’re all hot garbage. These companies have raised hundreds of millions of dollars and yet all look and feel like Kazaa but somehow even shadier. Honestly incredible.
If you believe healthcare crypto companies and DAOs eventually take off, then there are still a lot of tools to be built that can sell into them.
- How data storage actually works is a huge issue. Very abstractly, a blockchain can be thought of as many copies of a database that reach consensus with each other about the contents. Storing healthcare data on that chain is both not feasible because of how large the data is and not desirable for anyone that wants to keep that data private. Figuring out how to navigate on-chain and off-chain storage, making the data usable by crypto applications while maintaining privacy (e.g. maybe through zero-knowledge proofs), etc. seem like big problems.
- Tools to support healthcare DAOs are basically non-existent. There are general tools to help with things like gating DAOs, suggesting and voting on proposals for the DAOs to do, and collecting and dispensing funds. But there are probably going to be more things healthcare specific DAOs need. For example, Molecule** is making it easier for research DAOs like VitaDAO to allow members to invest and own IP in drug molecules the community is researching. What about a new form of Know Your Customer for people joining healthcare DAOs to verify their credentials but preserve their anonymity? Or how can data sharing agreements work for a DAO who may want to aggregate their members’ data for commercial purposes?
- There’s a whole set of off-the-shelf smart contracts to build that other crypto companies can use themselves. Insurance or pharmacy formularies are, at their core, a set of rules around what needs to be done in order to move money from point A to B. There’s potentially a future where someone builds smart contracts to encode those rules that other crypto companies can use or modify as needed.
I’m mostly talking out of my ass for this one, but I’m generally excited about the intersection of crypto x healthcare and looking for companies that can bring it closer to actually being useful.
Now you might be thinking to yourself…this is insane, there’s no way we’re anywhere close to crypto applications in healthcare. If that were true, then why did HHS put out a presentation about it with this deranged slide?
Most other industries have seen individuals build brands that are larger than the institutions they work for. Trust in healthcare organizations continues to fall while trust in doctors stays relatively flat. Clinicians really leaned into social media during COVID, for better or worse. People like Peter Attia, Eric Topol, Tiffany Moon, Austin Chiang, etc. all have hundreds of thousands of followers.
It seems inevitable that patients would explicitly want to see these doctors or use the products they endorse. If these clinician influencers wanted to start their own businesses and extend their brand to other places, how would they do it?
- Making it easy to set up their own practice, set up their accounting and banking, incorporate the practice, connect to their social media accounts to create funnels and drip campaigns, etc. Or after a while maybe even franchising the practice. Ease** powers a lot of this.
- Connecting their social media profiles to patient intake flows for enrollment into their clinics + content creation/marketing to the right cohorts.
- Creating branded products? Idk, a lot of my cookware is “a collaboration with Gordon Ramsay, because you have no idea what you’re doing”. Maybe there’s a microbiome supplement to be co-created with GI influencer, sunscreen from a derm doc, a calculator or some nerd shit from a nephrologist.
- Managing patient communities in a way that gets all the proper consents, stays HIPAA compliant, avoids malpractice if something gets construed as a diagnosis, etc.
- Financing to get set up and/or stabilize their salary compared to their current job while their patient panel builds.
- A crisis PR team for when they inevitably get cancelled for some terrible political opinion they let slip out.
There’s a whole list of “creator economy” tools to support one person businesses that can probably be adapted and/or made HIPAA compliant specifically for clinician creators that want to communicate and interact with patients in some capacity.
Also I hate the word “creator”, we all know there’s only one Creator (praise be).
Digital therapeutics seem to be finally hitting the clinics, getting reimbursed, and getting into patient’s hands. However, it’s still a relatively small category. Despite having 10,000 definitions of the term there are only about ~100 digital therapeutics products in the pipeline according to the Exits & Outcomes database. Companies like Pear and Akili have created a lot of the paths for second mover digital therapeutics to come to market in a much easier way.
Digital therapeutics are in some ways like pharma products or medical devices, in some ways like software, and in other ways like services. They’re unique in their own way and most have been trying to retrofit onto the rails of the existing healthcare system despite being totally new.
Some products where digital therapeutics could be the end customer:
- Helping them integrate into provider workflows and alerting them when a patient is a good fit for a digital therapeutic is still one of the bigger issues. Amalgam seems to have bought a clinical decisions support company to make it easier to surface digital therapeutics companies to providers.
- Having a salesforce that understands how to sell a digital therapeutic is tough because it’s different than selling a regular drug and requires more explanation. This seems to have been a point of tension for those that partnered with large pharma with the intention of using their commercialization muscle. Companies like Pangea seem to be offering turnkey sales forces to go to doctor’s offices which might work better.
- Decentralized trials are particularly useful for digital therapeutic companies who don’t really need physical sites. During COVID we saw relaxed rules around enrolling patients + setting up pragmatic trials with real-world patients, and I’d expect that to continue post-COVID. Companies like Medable and Curebase seem to already be powering these kinds of trials.
- Payment and contracting is a huge issue for digital therapeutics companies. Nisarg Patel wrote a great piece on this but in essence the digital therapeutic companies are setting up ad-hoc contracts with payers or they’re billing under existing CPT codes which aren’t really accounting for the device cost. Some companies have created digital formularies like Evernorth and Xealth to make it easier to prescribe + get reimbursed. But I think there’s still room to set up easy to use contracts for digital therapeutics vendors that track software usage to easily set-up value-based care contracts. Especially now as more payers start pushing back on digital therapeutics companies and demand more efficacy data from the real-world.
- As digital therapeutics start selling across different countries, maintaining regulatory compliance globally is going to become increasingly more complex. BrightInsight seems to have a product dedicated to doing this for pharma clients building their own digital therapeutic solutions.
For anyone building digital therapeutics companies, what’s a vendor you wish you had but didn’t seem to exist so you had to build it in-house?
This is once again me playing armchair analyst on the issues these customer segments face, but if you’re thinking about building a new company it might be worth talking to companies in these more nascent areas to hear their pain points.
What are some other totally new markets that you think are going to get big? Actually, let's just cut straight to the point. Send me all the companies you think are going to be really big, this investing thing is harder than I thought.
Nikhil Krishnan aka. “Wow did you just write one of those posts that tags a lot of companies for the express purpose of getting them to repost it? No.”
Other posts: outofpocket.health/posts
Thanks to Liam Mc Morrow, Morgan Cheatham, and Matt Schwartz for looking at drafts of this
If you’re enjoying the newsletter, do me a solid and shoot this over to a friend or healthcare slack channel and tell them to sign up. The line between unemployment and founder of a startup is traction and whether your parents believe you have a job.