On Restrictive Rules

2024.09.29

LXVI

I reflect on a sojourn into a McDojo and attempt to impart the startup related wisdom I learned here so you don’t have to waste your time recreating the adventure.

While it may feel that having more rules gives you more safety and comfort, it simply makes you more fragile. You become stronger through exposure to randomness and chaos, not by hiding from it.

The fewer rules a martial art has, the closer it is to a street fight, and the more equipped it is to deal with both the real world and other martial arts that follow more rules than it does. 

Likewise, it seems that the best way to run a startup is as if you were in the middle of a street fight all the time. If you act like there are strict rules that must be followed, you’ll be in for a surprise when you get kicked in the you-know-where. 

McDojos

I went to a martial arts class at a college in Boston*. It was a fifty something year old man throwing around four un athletic 20 somethings. He had them punch each other in slow motion and then do some choreographed stunts. We’ll call this martial art Mickey Mouse Self Defense.

A good litmus test for the robustness of a martial art is whether or not it involves live training. When I asked if they do anything of the sort, the Sensei said no. Of course, Ronald McSensei told me that my Brazilian Jiu Jitsu training would be an impediment to me understanding his martial art. Moreover, he had moves that were so lethal he was not even able to demonstrate them for me!

“Martial Arts” like this exist in a fantasy world. They tend to have a lot of rigid rules, unrealistic expectations, and little action. On the other hand, something like MMA has few rules. It is more “extensible” and exposed to reality. For this reason, I am far more afraid of someone who trains MMA than I am someone who trains in Mickey Mouse Self Defense or wrestling or even Brazilian Jiu Jitsu.

*Leaving both the name of the martial art and the particular university anon to hide the man’s identity. Ifykyk.

Extensibility

In Brazilian Jiu Jitsu, the game goes until you or your opponent taps out. There aren’t a whole lot of rules stopping you from doing many things along the way. Interestingly enough, you can even recreate just about any series of moves from either wrestling or judo*. And, you can do even more than you can in either of those sports. In that sense, BJJ is a “superset” of both wrestling and judo.

Someone who does BJJ will gladly grapple with someone who does wrestling, judo, or any other non striking martial art. BJJ’s lack of rules make it robust to these “outside techniques.” After all, these outside techniques are contained within BJJ itself.

If I as a BJJ practitioner were to do an MMA match, however, I would not stand a chance. The biggest constraint in BJJ is that there is no striking. MMA is like BJJ minus this rule. In that way, MMA is a superset of BJJ, and wrestling and judo, and Muay Thai and boxing and kickboxing and even Mickey Mouse Self Defense.

MMA has the most free variables. It is the most expansive martial art. But, do you know what has even more free variables than MMA? 

A street fight. 

*While “spiking” and Kani Basami are illegal techniques in BJJ that come from wrestling and judo, these techniques also tend to be illegal in their sports of origin, even at the highest levels.

Startups

With startups, the more you act like there are artificial constraints holding you back, the slower you go.

I’ve tried following the “instructions”. Come up with an idea, find out how long it will take to build, write three or four pages on why you have a moat, make a financial model, a slide deck, incorporate, etc. Unless your product needs serious upfront capital investment, this is the Mickey Mouse Self Defense of startups–a collection of rules that give you a false sense of safety and security. “Do these things and it will work.”

From experience, those things do not inherently work.

While I’m not qualified to tell you with certainty what does work yet*, I am finding that the MMA version of startups seems to not only involve far fewer rules, but it seems to be much more effective. Build a good product and get people to use it.

*Remember, I am still living with my parents

Open Source

A startup that is showing all signs of being on a strong trajectory is OpenBB. This is not surprising; the founder, Didier, ships like a mad man. 

What is surprising to me is that they open sourced their technology in a space where such an idea seems absurd–they operate in the financial data space. Open source feels orthogonal to the secrecy you would expect in this space.

If startups are MMA, this is the equivalent of turning it into a street fight. OpenBB was effectively able to benefit directly from the engineering ability of anyone interested in using the technology. This is powerful leverage when the incumbent in the space has 11 figures in revenue. 

Maximizing Surface Area

While we will not be open sourcing BirdDog, we will be taking a page from the OpenBB book to increase the surface area on which we’re receiving feedback. 

Our product let’s users define data points they want to observe across their prospect list by asking questions. On the backend, this requires us to rework the question, select the return data type, add some keywords, and maybe fine tune it with an example. This is a choke point in our feedback loop. We have our own assumptions about how the technology we built should be used. 

Still, it is not at all obvious that we should make these other variables accessible to the user. And, really, it is probably a “bad idea”. Our users are typically sales people who want complexity abstracted away. 

That being said, some of our users are a bit more tech savvy and would love to try playing with these variable themselves. Letting them do so would increase our rate of feedback.

So, it seems we’re at a crossroads. Either we hide the variables and make the product feel like a service, or we open the variables to increase the rate of feedback. 

Really, thinking in either ors right now would be the startup equivalent of Mickey Mouse martial arts. The service route would probably be wrestling, and letting the user edit the variables would be maybe BJJ or MMA. 

Selectively revealing the variables to the users who are interested in them or who we think would feel comfortable editing them, and letting them do so directly, however, feels the most like a street fight.

We have no idea if it will work or not, but the asymmetry in being so early is that it doesn’t really matter if we’re wrong. Given the lack of evidence to prove that we’re “right” about anything, we’re already “default wrong.” So, we’re going to try it out.

In summary, rules represent assumptions about the real world that are exempt from further testing in an environment. 

I think the key is to remember that all assumptions are always testable.

Live Deeply,