The Fundamental Law of Startup Success
Welcome to Weekly Roundup #18 at ForecastOS! We're glad to have you here, following our journey :)
Over the past week, I've thought a lot about startup characteristics like:
- vision,
- product-market fit,
- execution / quality of work completed,
- iteration speed,
and how these characteristics contribute to a startup's likelihood of success.
When thinking about what drives startup success likelihood, I'm reminded (as any REAL quant investment geek is) of the fundamental law of active portfolio management:
The Fundamental Law of Active Portfolio Management
information ratio (IR) = information coefficient (IC) * transfer coefficient (TC) * √breadth (BR)
The fundamental law of active portfolio management tells us that investment performance is a function of how often you are right (your information coefficient, IC), how efficiently you can translate your forecasted returns / outcomes into portfolio weights / positions (your transfer coefficient, TC), and how many bets you can make (breadth, BR).
A casino, for instance, has a relatively low IC (barely above 0), a very high TC (1, in unconstrained games), and VERY HIGH breadth. A casino, like a traditional quant investor, makes lots of bets, and makes money through quantity of bets (vs quality of bets).
A similar law can be applied to startups if you change the variables and formula slightly.
The Fundamental Law of Startup Success
A startup's probability of success (S) is a function of how often the founders' vision has product-market fit (P), how often a startup can execute (i.e. build and market) its vision (called transfer coefficient, TC, in investing, called execution, E, here), and how many product launch iterations a startup has (L).
Probability of Failure (F) = (1 - P * E) ** L
Probability of (at least 1) Success (S) = 1 - F = 1 - (1 - P * E) ** L
For example, let's look at 4 contrived examples with made-up success likelihoods.
Applying The Fundamental Law of Startup Success
Example 1: the industry-insider
This could represent: industry insiders that have strong product vision (i.e. vision has a 50% chance of PMF), but that can't build / execute effectively, and only launch 1 product.
P = 50%
E = 10%
L = 1
S = 1 - (1 - 0.5 * 0.1) ** 1 = 5.0% chance of success
Example 2: the hungry undergrad
This could represent: fresh, hungry undergrads that are trying to disrupt a space they don't understand well (vision has PMF 20% of the time), but that execute fairly well (33% of the time) and get a couple iterations in (2).
P = 20%
E = 33%
L = 2
S = 1 - (1 - 0.2 * 0.33) ** 2 = 12.8% chance of success
Example 3: the cockroach
This could represent: founders that manage their runway extremely well and refuse to let their startup die. They may not understand their space well (vision has PMF 20% of the time) or execute well (successful execution 20% of the time), but they stay alive long enough to get 5 big swings at the proverbial product plate.
P = 20%
E = 20%
L = 5
S = 1 - (1 - 0.2 * 0.2) ** 5 = 18.5% chance of success
Example 4: the hungry, frugal, skilled industry-insider
This could represent: motivated founders that understand their problem space well and have experience building end-to-end solutions quickly. They have strong vision (50% chance of PMF), execution (50% chance of execution), and velocity (5 launches).
P = 50%
E = 50%
L = 5
S = 1 - (1 - 0.5 * 0.5) ** 5 = 76.3% chance of success
tl;dr: The Fundamental Law of Startup Success
Understand the problem, build the right solution, launch the solution to users and collect feedback, and repeat. As fast as possible.
Learn, build, launch, repeat. Faster. Faster! FASTER!!!
There is no panacea.
Work: What's Coming Next
As was the case last week, we are working on our out-of-the-box quant investment strategy and our first fully-owned dataset, soon to be available through ForecastOS FeatureHub. They are both targeting an end of March completion date. We can't wait to share them with you then!
Until next week,
Charlie