Why Arivale failed: Behind the scenes of the first venture-backed medicine 3.0 company

Over the next five years, I predict we will see ten unicorns that focus on consumer longevity. Medicine 3.0 is about to go mainstream.

But what most people don’t know is that in 2015 a little-known company called Arivale already attempted this, raising over $50 million in funding, hiring 120 employees, and acquiring 5,000 customers for diagnostic testing and coaching.

Four years later, Arivale failed.

Many friends are now exploring ‘Arivale 2.0’, so I decided to write a few points for them and every other health optimizer turned entrepreneur who thinks Arivale 2.0 is an exciting idea.

But first, what exactly was Arivale?

Arivale was a consumer healthcare company that provided ‘scientific wellness as a service’. For $2,000 per year, Arivale would test your genetic makeup, blood, saliva and stool, then pair you with a wellness coach.

But that was just step one. Behind the scenes, Arivale had a second business model: gathering data to predict early disease onset.

A bold mission, that’s for sure. But Arivale presumably had the team for it, headed by famed geneticist Leroy Hood who has founded 15 biotech companies, has 18 degrees, and published more than 850 peer-reviewed articles.

Key reasons for Arivale’s failure

• High CAC: Arivale struggled to efficiently acquire customers. This was partly due to inefficient marketing spend: it turns out events and parties don’t drive acquisition as well as performance marketing. Arivale’s team didn’t have consumer GTM experience and this showed up in their exploding CACs. The best category creation playbooks I’ve seen in consumer health focus on (a) content-led growth to educate consumers on the category and (b) influencer-led growth to change culture. Loosely, education + culture change = category creation. Category creators often market the category as well as the product. Think Levels with metabolic health, Liquid Death with canned water, Oatly with oat milk, and Tinder with online dating.

• Small market: Arivale’s CEO said “I do not believe that there is a meaningful market for a program that’s helping people do something in the future.” I agree, but I think he misses the forest for the trees: good startups create the market. Failure to create a market is either a (i) failure of execution, or (ii) lack of first-principles reason why the product should exist. In this case, I think (1) was more true than (ii).

• High clinician costs: Arivale hired too many doctors and had low utilization. Doctors are expensive. Improving utilization is one of the primary ways to reduce COGS for a tech-enabled services company. High ARPU can make companies like Arivale complacent about COGS optimization. That’s okay in the short term if you have strong retention. But slim gross margins alongside a weak retention profile is a failure waiting to happen.

• Poor retention: At $3,500 a year (later revised down to $99/month), you need to provide a lot of value to retain patients. Arivale’s challenge was that the benefits of the testing were front-loaded. This is because the genomic insights were more useful than other omic insights, and your genome only needs to be tested once.

• High cost of testing: Deep phenotyping is expensive. Arivale’s thesis was that the cost of multi-omic assays would come down, much as genomic sequencing costs have reduced faster than Moore’s law ($1 billion in 2000 to $100 in 2022). However, cost reductions didn’t happen fast enough. As a result, Arivale ran at a unit economic loss which means that Arivale loses more money the bigger it gets.

• Unwillingness to pivot: The Arivale team was steadfast in pursuing their high-cost flagship program, even when market data suggested otherwise. They were slow to pivot on the GTM strategy when their existing channels weren’t delivering positive ROI.

Key product takeaways from Arivale’s program

• Be wary of gimmicky insights: Genetic insights felt gimmicky (and were gimmicky). Same with microbiomic insights (or frankly, lack thereof). Patients see through rhetoric, so be warned consumer healthcare wrappers – you need to solve real problems, not just make something sexy and cool (cough Forward).

• Splitting health into categories is helpful: Arivale split health into five dimensions: heart health, diabetes risk, inflammation, optimal nutrition, and stress management. I like this product choice because it reflects how patients think about the human body.

• Health coaching is tough to quality control: Arivale’s program was too dependent on the quality of the health coach. But novel uses of AI and rules-based systems can start to change this.

• Human interaction with the health coach drives outcomes: Human accountability is powerful. There’s lots of evidence of this from companies like Livongo, Future, Omada, and Noom. Contact me and I can share my notes on those companies.

• Breaking weight gain into four ‘causes’ is neat: mental (anxiety, boredom, dopamine), genetic, metabolic, and hormonal. Some companies implicitly target part of this stack. Levels addresses metabolic causes of weight. Noom addresses mental causes of weight. GLP1s address hormonal and metabolic causes of weight.

• Don’t set target numbers: An interesting behavioral choice is that Arivale told members not to set target numbers “Research shows numbers are not as motivating as what we think they should be; they’re ultimately not what makes changes for people.”

• Coaching is onerous for consumers: Arivale required members to track their weight, monitor their water intake, and upload photos of food into the app. The problem with this is that consumers care more about “make sure nothing is wrong with me” and “give me actionable insights” than “give me coaching”. I haven’t seen many companies succeed that combine coaching with diagnostics/insights.

• Be careful of generic insights: After each session, Arivale would send a summary of the topics discussed. But customers found that the summaries felt generic, as if they were from WebMD. It’s tricky to strike the balance between providing education and not overwhelming customers.

Is Arivale 2.0 possible?

I think Arivale 2.0 is inevitable for several reasons:

1. Longevity is about to go mainstream. The likes of Huberman and Attia are growing in popularity. There are more founders targeting this space than ever before. And consumers want more from their health.

2. The cost of testing is rapidly reducing. What was once available to the few will soon be available to the many. This is the inevitable arc of technological progress.

3. The savviest companies in this space now understand insurance reimbursement. This makes it possible to provide an Arivale-esque service for ~free, which will drive demand.

But there are several challenges to building in the space:

1. Direct-to-consumer customer acquisition is expensive and challenging. Companies with existing customers have an advantage. As do companies that have higher LTV products to outbid others on performance marketing.

2. Margins on diagnostics will trend towards zero which means that companies will have to make margin elsewhere. We will see lots of companies make pricing opaque by bundling diagnostics into a larger program.

3. Market size is still small. The ratio of people interested in health optimization to people building in health optimization is too high. In other words, lots of people interested in this space are also building companies in this space. There will be a lot of people building niche products for their friends in Silicon Valley that never manage to go mainstream

If you’re exploring starting an Arivale 2.0, feel free to get in touch!