Year 2 Startup Learnings

5 October 2021

I incorporated Loyal in October 2019. Last year, I wrote my Year 1 Startup Learnings.

Since then, we’ve gone from 12 people and a lead drug candidate to a team of 30+, two IND-stage drugs, our first formulated product, and our first clinical study successfully completed. While I could write (many) essays on the technical knowledge I’ve gained in the last year, this will focus on more meta learnings.

You do not need to fully visualize the endpoint to realize it will be valuable.

One of the better decisions made this year was building out a native engineering & data science team at Loyal. You do not traditionally need much engineering talent to bring a drug from idea to market approval. The core reasoning was that smart people with different backgrounds running together on a very challenging problem usually produces good things - there was no specific workstream or goal more concrete than that. This became much more true than I initially predicted: this team has already saved us 7-figures+ in potential wasted efforts, and materially reduced the probability that our drug products will fail for reasons unrelated to the biology. [1]

This also reinforced the criticality of mindset/thinking diversity within teams. The engineering culture and thinking frameworks have been a fantastic augmentation and counterbalance to the more traditional scientist and clinical mindsets, as have the scientist’s frameworks to the engineers.

Functionally, this looked like hiring someone with strong technical competence who was excited to build their career in aging with me at Loyal, and then giving them free rein and resources to run on their vision of a core computational competency at an early-stage biotech company.

As you get more information, previously good decisions can turn bad.

One of the more common failure modes at Loyal is decision creep: Given [X, Y, Z], we decide something. Over time, these variables shift to [X+1, Y, Z], then [X+1, Y-2, Z], etc. No single shift is independently notable, but at a certain point the cumulative movement reaches critical mass and prior decision(s) transition from correct to incorrect.

Catching this is difficult; we still haven’t mastered it. You do not want to/cannot reassess every decision and variable or you will never get anything done, and many variables are impossible to predict with certainty anyhow until you just do the thing. However, you clearly want to minimize avoidable failures and maximize probability of success. We have learned that recording decisions and the reasons behind them is very important, as well as re-assessing decisions and key assumptions at regular intervals.

You will feel less.

Seemingly an adaptive responsive to stress, over time the highs don’t feel as high and the lows don’t feel as low. This is net helpful - you do not over-index on the positive news and you do not crash when things go wrong. Steadfastness is a key leadership trait. Unfortunately, emotional variance is a key human trait and a flattening affect can make true ‘happiness’ seem unattainable.

I have had to put a non-negligible amount of energy into resisting indexing on false-positives in an effort to ‘feel something’. For the first time in my life, a small subset of people know who I am before I introduce myself, my (and Wolfie’s) face has been shared in some of the best publications in the world, and - finally - I am no longer in the bowels of the Silicon Valley social hierarchy. [2] This feels nice, but none of it matters. It’s tempting to lean into this stuff because it satisfies some reptilian, shallow desire and is an easy way to break out of the flat affect induced on you by startups. You have to resist!

With scale comes increased ability to help others.

The decision I am proudest of this year was using my platform to bring to light a predator in the aging field. [3]

Because of Loyal, I knew I could come forward without destroying my career; this would have been untrue up until now. I was privileged to be able to represent the many other women who have suffered at the hands of Aubrey, most of whom did not feel they could do so themselves.

[1] Sorry I can’t be more specific! One day.

[2] I wrote a bit about learning how to fit the pattern here. It’s positioned towards women, but should be broadly relevant.

[3] You can read my story here.