prioritizing the right activity inside your company to win

Everything is a system

Everything from engineering and biology to filmmaking and democracy is a system — each with its own moving parts, all interacting with the world outside it.

Some, like calculus or thermodynamics, have well-defined rules; others, like building startups, discovering new math theories, or solving new engineering problems, are far more dynamic.

Capitalism in its current form disproportionately rewards people who are good at dominating the dynamic systems. The number one skill required to succeed is to mentally model that complex reality and take a profitable position inside it — understanding the components, how they interact, and developing an intuition for what winning looks like.

Civilizational technologies like AI challenge the current assumptions and rules of almost every system. The nature of the system changes, and thereby so do the activities and what it rewards. As AI gets better at understanding and dominating systems with pre-defined rules, the value of human work inside a dynamic system will drastically go up.

Let’s take the example of computer science. In the 2000s, writing good code was hard and time-consuming as there were few pre-determined rules, but with every framework abstraction or open-source contribution, engineers understood what good code was. Today AI has completely replaced writing code as the rules are much more defined.

At the same time, new sets of unstructured problems are emerging and are becoming valuable — inference engineering, system design, MLOps, AI research, evals, etc.

Building a startup is one such complex system, and recognizing that can liberate you from the unstructuredness and uncertainty, and put you in the right frame of mind.

Activities to create winning conditions

Every system can be further broken down into sub-systems with their own set of rules, assumptions and activities. For example, a big system like math is made up of sub-systems like arithmetic, algebra, calculus and more.

These sub-systems interact with each other and also with external systems like physics, engineering, or finance.

A startup is no different. The usual sub-systems are product, engineering, R&D, marketing, sales, hiring, and fundraising.

The founder’s role is to mentally model the reality of the environment they’re operating in, set a winning strategy, and determine how activities get carried out every day in achieving it.

As the title of the essay says, I will be focusing on the last part — to prioritize the right activities¹ carried out by the sub-systems to win.

The reason to look at these as sub-systems is not to create rigid boundaries, but to have a map to navigate them effectively.

Define what winning looks like for each sub-system

• Write down the good, the bad, and the ugly: define these outcomes clearly, and make sure everyone involved understands them. The good is the target, the bad is an opportunity to give feedback, and the ugly is a signal to intervene immediately. 
• Each part should amplify the whole: the values and impact of the company must be amplified by each sub-system. Everyone should model winning behaviour in a particular environment. For example, in a prosumer category, it means constantly shipping content-led launches; in B2B, it means owning the relationship and feeding insights back to the product.
• Founders / leaders should be hands-on: each sub-system must be deeply understood by the founders, who need a clear mental model of how it all fits together. If one sub-system is constantly falling behind, it’s time for the founders / leaders to get their hands dirty and solve the issues.
• Observe and fix silos: unless a sub-system handles something legally or competitively sensitive, such as fundraising terms, there’s no reason for silos. Identify information bottlenecks or interpersonal issues and fix them early on.

Fewer activities, but done well

Instead of doing too many things or being pulled in too many directions, it is better to stick to a few activities to deliver on the winning strategy.

With few things at hand, each activity can be done well, and every sub-system reinforces the others, strengthening the company’s structural position.

Each activity should have a clear owner and a well-defined scope of work.

Hiring the right people to sustain the activities at the highest level possible

Hiring is a horizontal sub-system, and its quality sets the ceiling for every other sub-system.

The golden rule of hiring is: the right person, in the right place. Be it an intern, a leader, or a co-founder.

There are many layers to hiring the right person, but they can be categorized into three broad categories:

• Competence: performing the craft very well, at the highest level possible.
• Cultural fit: resonating with the company’s values, and exemplifying and amplifying them.
• Slope: how fast they can take on more responsibility as the company scales.

Every step of the hiring process should be to collect as many data points as possible for the above three layers.

I like Jensen Huang’s approach here: hire so rigorously that firing becomes unnecessary.

Every person hired should clearly understand what winning looks like inside their sub-system. The incentives should be aligned with winning.

Some activities don’t have immediate ROI, but can become highly valuable / disruptive later

Lead generation or sales operate at much shorter time horizons, while good, quality R&D takes time but can have a disruptive impact on the business and markets.

Different sub-systems of the company run on different time horizons, so founders need to keep them aligned for real impact.

For the ones with a longer time horizon, founders and leaders need to be more hands-on, since they have more context about the business and the markets. The activities must be monitored and fine-tuned with tighter feedback loops.

In conclusion

I wrote this essay as a way to clarify my own thinking on how to streamline activities inside my startup. I hope it also helps you and acts as a guide for you in streamlining yours.

Footnotes

1. I chose the word activities to reflect the thinking of value chain analysis by Michael E. Porter. ↩