I have thought about scaling before, as an orthogonal problem to whatever the original problem is. Scaling a website is quite different from building a website. More generally, we can think about problem solving skills and their higher order versions in a similar fashion. In this lens, scaling is a broad category of such examples, but not the only one.

Suppose that being a good coder (i.e. able to produce good code alone) is a first order skill. A higher order version of the skill is being able to produce good code in a group. It is higher order because it’s an extension of the original, first order skill: knowing how to write good code solo is a necessary but not sufficient prerequisite. When other people are involved, things that were not as important (such as documentation, communication, social skills) become critical, in addition to whatever is needed to write good code.

An even higher order skill is being able to create groups of people that can produce good code. Here, we are elevated from the position of an individual contributor within a group to a group-level creator. Not only does one need to know how to write good code and how to work effectively in a group, but also how to create environments where productive groups can form and thrive.

Back to being a good coder as the first order skill. From a different angle, another higher order version of the skill is being able to hire other good coders. It may sound fairly straightforward if you are already good at something yourself, but it doesn’t quite translate. To be good at something does not necessarily require an understanding of why you are good at it, however hiring does. In fact, hiring also benefits immensely from knowing what one’s own weaknesses are. This is an area that many people (especially startup founders) miss. They will do the first round of hiring themselves, which tend to work out well (i.e. the people are good at their jobs). But then the new people are entrusted with hiring the next set of people, which don’t always work out, because they are not good at hiring despite being good at the role themselves. This is a case of failing to recognize the difference between the two different orders of skills.

Despite sounding fancier, higher order versions of skills are not strictly superior, and not everyone should be pursuing them. It is perfectly fine to focus on honing one’s skills at a particular meta-skill level, with no need to go higher. After all, the problems faced at each level change, and people all have their own preferences on what they enjoy solving. However, the process of scaling (e.g. of companies or organizations) often demands higher and higher order skills of the people involved. So people who start companies on their own or are in management positions during organizational expansions have no choice but to develop these skills.

There are no silver bullet here to acquire higher level versions of skills, but the hiring example above illustrates a critical point on the importance of introspection. Since higher order versions of skills are typically meta to the original versions, being able to recognize and fully understand what makes an individual good or bad at the skill is extremely crucial. To do this, assuming one is already good at the original skill, requires deep and genuinely unreserved reflection. Recognizing and completely internalizing one’s own strengths and weaknesses around a particular skill is often unpleasant, which is why many avoid it, but it’s necessary in order to improve at the higher order version of the skill.