Capital matters when you build hard things

3 April 2022

Loyal recently crossed $50M raised to develop drugs intended to extend dog lifespan and healthspan.

A common refrain is that fundraising and headcount are not real company milestones. While technically correct - the milestones come from what the team does with the money - I disagree with the behavior this encourages: founders & teams not celebrating fundraises and not celebrating building their team. It is hard to raise money and build a team, even more-so when you are working on a problem outside of the well-trodden Sand Hill Road paths and patterns. Being well - not just sufficiently - resourced is also key to tackling a technical problem well.

Many of the most relevant problems facing humanity today are technically challenging, and many breakthroughs are lost: not because they are impossible, but because of some combination of bad luck on the initial bets made and running out of money.

More importantly - and the point of this post - is that capital itself is strategically critical for companies working on hard, deeply technical problems.

1 - The team you need, needs stability

When building a startup in a deeply technical field, you must walk the line of respecting the old guard, while pushing on their priors and biases. Especially in scientific disciplines, you cannot figure out everything from first principles - there is an undeniable necessity of the context only earned from decades of work. As the founder, you need find and hire these people as soon as you can.

Often, these types of hires are at a different stage of life. Common tech stability signals such as working with a top fund will not carry weight with this type of hire. They often cannot or will not work for pennies; even if they want to, they often have responsibilities to mortgages, college funds, and savings. You both need to be able to pay them enough and have a degree of company stability.

An example:

We hit our key Series A milestones at Loyal with less than $500,000 of our original $5M seed raise. Sometimes I feel frustrated about selling so much of Loyal at a seed valuation. It is however unlikely that I could have built the team Loyal needed to without that $5M in the bank - it provided enough stability that my early team felt like they could both take a risk on Loyal and not betray their responsibilities to their family.

2 - You can experiment in parallel

When working on hard problems, no matter how good your team is, you will need to conduct experiments where you are unsure of the results. In less well funded situations, you’ll often iterate in sequence - do experiment 1, take your learnings from that and do experiment 2, etc. Unfortunately, experiments in most technical fields have a significant time variable and you can burn years experimenting in sequence.

Capital allows you to parallelize experimentation, front-loading your learning and allowing you to cycle much more quickly (at the expense of your bank account).

An example:

One of our products is complex - a dissolvable implant. Implants are challenging because of the number of variables you need to tune to get the desirable product (rate of dissolution, dose of the drug, pharmacokinetics of the drug over time, COGS, manufacturability, etc). Early in Loyal, we had to experiment in sequence - we would make one variation, see why it failed, make another variation tuning from what we learned, and onwards. While this meant that each cycle was more informed, it also meant that we were anchored to our first “guess” at the best formulation. Now that we have more capital, we do a lot of work like this in parallel. We then can take all of the learnings and have a much better shot on goal on the second cycle.

3 - Run a cleaner experiment & build in resiliency to (bad) luck

Most deep tech companies are built around a core thesis or insight (“company bet”) that is then leveraged to build a better/cheaper/new product. Validating that the company bet is correct often is very technically challenging, expensive, and time intensive.

It is often difficult to run a clean experiment when working on technical problems. “Clean” here means that you are testing the company bet without confounding the test with other variables. Confounding variables mean that your shot on goal (the product validating the company bet) may fail despite being the underlying bet being correct.

Using the example of Moderna:
Core company bet: mRNA-LNP technology can provide a safe, effective vaccine against infectious disease
Shot on goal: mRNA-LNP for COVID-19
Potential confounding variables: specific biology of COVID-19 and its relative lack of understanding at the time

You can insulate your company from false fails by testing multiple variations of your company bet in near parallel. While each shot on goal will still be vulnerable to failure (due to an incorrect bet or confounding variables), it is less likely that you come to a false conclusion - that your company bet is wrong - if it is actually correct, assuming you choose your shots on goal well.

Doing this well - you guessed it - takes significant capital.

A couple of final notes

To me, the idea of a good technology failing for reasons unrelated to its viability is terrifying. False failure burns the area for future founders, and can mislead and delay necessary progress. More capital going into companies and people working on hard problems can limit this failure mode.

In addition to helping build Loyal with the highest probability of success, this $50M raised milestone means a lot to me personally. There are not many role models for young women wanting to build deeply technical companies and play in the sandbox with the boys. I am incredibly thankful for those who have backed me - monetarily and personally - and who have given my team and I a fair shot at building something important.

A higher probability of succeeding on our audacious goal of canine life extension - that’s something worth celebrating.