3 Mistakes (and Lessons) Building a Digital Health Company
Reflections Part 1
Over the past few months, I began reflecting on my past four years building Songbird.
Reflecting on a failure is difficult. You start a company because you have an optimistic vision for the future and high expectations for what you can achieve. Failure brings that crashing down. The exciting future you saw is merely a distant mirage dissolving into nothing. Failure also deals a powerful blow to your identity, making you question who you thought you were. Am I cut out for entrepreneurship? Were my past successes just luck? When I reflect on my mistakes and experiences, certain fears come out of the shadows and add to the difficult emotions.
Over time, though, it gets easier. Space and distance cause self-critical thoughts to fade. Mistakes and experiences begin to feel less charged. A past decision isn’t stupid, it simply… is. And I begin to glimpse a reality that is no longer so tinted. From this emerges something useful: learning.
Reflection is fundamentally optimistic. To reflect is to believe that there is a future worth reflecting for. There will be a next time, and next time will be better.
Below are the top three mistakes (and resulting lessons) from my journey building Songbird. Writing this was helpful for me, and I hope these lessons can be helpful for you.
Mistake 1: Launching and Growing in the Wrong Geographic Markets
We launched in the Bay Area in 2020, but ultimately discovered it was the most difficult geography to build an ABA business. The gross margin economics behind ABA are simple: gross profit of 1 hour of care = reimbursement rate - therapist wages1. With high gross margins, you have room to work with, and with low gross margins, you’re scrounging to be break-even. In the Bay Area our gross margin was capped at ~25%2 because of mediocre reimbursement rates (~$55) and high wage costs (~$25-30).
When your gross margin is slim, you need a much larger revenue base to justify the same amount of OpEx spend. But the same underlying cause of the poor gross margin, high cost of living, also led to intense labor supply shortages and retention issues. We could always find demand, but couldn’t scale supply, and therefore struggled to grow. We eventually found ourselves with most of our revenue in the Bay Area and a low-margin subscale business.
Andy Rachleff, formerly of Benchmark Capital, said it best: “When a great team meets a lousy market, market wins. When a lousy team meets a great market, market wins. When a great team meets a great market, something special happens.” He refers to the broader market category one is in, but the principle extends to geographic markets as well.
There are two moments where this mistake occurred. The first is when we were evaluating our initial markets. We let the location of a quality clinical hire dictate where we launched when we should have run a rigorous market selection process. The second moment is at around ~$1m run rate into a market when we could have terminated or stopped growing.
The Lesson: You want your business to be in a great market. A rising tide lifts all boats, and a stormy sea litters the beach with driftwood. Know which choices are the impactful irreversible decisions that asymmetrically contribute to success.
Mistake 2: Financial Discipline
As a tech-enabled healthcare service business, financial discipline needs to be a key competency. Financial discipline means ensuring the dollars you spend (headcount, marketing, new initiatives) are giving you the appropriate return (growth, profit, capabilities). We made two critical mistakes here: managing gross margin as well as burn multiple.
We mistakenly oriented around gross revenue when we should have been focusing on gross profit. Bessemer has a great article about how tech-enabled healthcare valuations should be on gross profit versus gross revenue. I agree. We had negative or near-zero gross margin for too long because we were underutilizing our salaried clinicians. (We overhired clinicians in advance of growth, eating some of their salary when we didn’t have the patients to fill their caseload). This strategy was a mistake. It may be reasonable to be unprofitable as a tech-enabled services business, but there are few reasons why gross margin should be negative.
Our second error in financial discipline was on burn multiple. Burn Multiple = Net Burn / Net New ARR. At its core, burn multiple implies that you should be getting something from burning cash. It’s best if that something is growth, but it can be capabilities as well e.g. R&D, payor contracts. Any burn should be critically examined… let’s not forget that most businesses don’t raise venture capital and always operate profitably, and many of those can still grow at meaningful rates!
Our burn was too high, and was not justified by our growth rate. Burn multiple would have helped us catch and fix the previously mentioned market selection errors and mismanagement of clinician utilization. Moreover, further below gross margin on our P&L, we had overhired headcount, spent too much on marketing, and invested in non-critical initiatives. Critically evaluating burn multiple would have helped us question these practices.
Poor financial discipline led us to spend our VC cash faster than we should have. But it fosters a lack of operational discipline that is even more insidious. When you allow negative gross margin because you’re focusing on growth, you allow for poor demand/supply scheduling processes. When you allow for high CAC because you overhired salespeople, you enable a culture of low productivity. And in an operationally and people-intensive organization, these norms are like ink in water, and are hard to clean up later.
The Lesson: “You are what your record says you are.” Your P&L is the only true measure of your business, so manage it carefully. Financial discipline breeds operational discipline.
Mistake 3: Failing to Identify a Venture-Backable Opportunity in Autism Care
We raised venture capital to reinvent autism care, but eventually found ourselves running an in-home healthcare service business. We had built a real business across 3 states, but it did not have the potential to grow into something much bigger on the right timescale.
One way I think about it is we didn’t find product-market-fit in our industry. People are surprised when I say this, because we had many millions of revenue and customers who loved our service. But in healthcare, product-market-fit requires that you create unique value by capitalizing on a “leverage point.” A leverage point might be software, virtualizing care, or differentiating one’s clinical model and using that to achieve payment model innovation. Leverage points are necessary to have the possibility of building something venture scale3. We had achieved “service-market-fit,” but as I heard from another tech-enabled service founder in the insurance space: “it doesn’t mean much to say you have PMF in the market for corn by selling corn.”
There’s nothing wrong with building a services business. I believe ABA can—in the right geographies, with the right discipline—be a fine business. However, the moment we raised $10 million of venture capital, we created a path dependency that necessitated venture-scale outcomes.
The Lesson: Don’t mistake topline growth for product-market-fit. Have the patience to take your time early on, because the insights that enable large and unique market opportunities take time to form. Be realistic about whether you are building a venture-backable business. Most business opportunities aren’t, and that’s fine. It’s easier to raise venture later on to capitalize on an amazing opportunity than it is to go backwards and change your preference stack.
True costs to calculate gross margin should use fully-loaded wage and be adjusted for non-billable activities, but reimbursement - wage is still a good approximation. Fully-loaded wage includes taxes, benefits, and workers comp. Non-billable activities include admin time, drive time, cancellation pay, etc. which always exist even at strong operational efficiency.
~25% was the limit of possible gross margin assuming our operational metrics (BCBA utilization, cancellations, etc.) were best-in-class. But if you’re trying to grow, it’s very difficult to be running the machine at peak efficiency.
There may be a more accurate term other than “leverage point.” My point is that the market structure of healthcare requires that digital health opportunities be evaluated on their viability to be venture scale. Software (not all, but some) can be venture scale because the marginal cost to produce is low. Many services businesses are not venture scale, but some can be - e.g. telehealth services are easier to scale and if clinical innovation is involved then the business may be able to build a moat with payors via specialized contracting (Cityblock for dual-eligibles, Equip for eating disorders)
thanks for sharing these lessons! They're useful for me to learn from.