The Infamous 2%

That is bound to be the most misleading title in the humble history of this blog. The 2% I am referring to are not related to the much more famous 1% of Occupy fame. They represent the standard allotment of shares given to the technical leader of an Internet startup.

In case you didn’t know, there is such a thing as a standard setup for a venture-backed software startup. Basically, the founders share according to contribution in roughly equivalent parts. People that join later (where later is loosely defined as “after the first funding event”) get a much smaller part of the pie.

In particular, a newly hired CEO get 10% of the company, while the CTO/VP Engineering gets 2%. Going against the grain of this wisdom is anathema to a VC, and if your company deviates too much from this scheme, it draws unwanted scrutiny.

The reverse also happens, and I have been asked several times to join a startup as founding developer and head of development, only to be offered no pay and 2% of the company. The deals I witnessed were particularly bad: once, a guy had a patent filing in hand and no intention to do much of anything for the new company, demanding I quit my job for the opportunity of getting 2% of the pie. Another time, a group of technically incompetent people had a vague idea of what they wanted, and thought that the guy that was going to make it all happen and shoulder the work was not worth more than one fiftieth of the company.

At the same time, I note that the most successful Internet companies, with few exceptions, have technical founders that stayed on to lead the growing company. This is true for Google most of all, of course; but Yahoo worked the same way; Facebook’s Zuckerberg is known as lead hacker from the movie, The Social Network; Dorsey of Twitter fame used to be a web developer. Most remarkably, compare the Microsoft of uber-hacker Bill Gates with that of his uber-MBA successor, Steve “Chair Man” Ballmer.

There is a part of this that is simply criticism of the Venture Capital system: VCs are uncomfortable with hackers and developers because there is just no personal connection between the two. My sense has always been that VCs choose whom they want to interact with at least in part based on how much they’d like to sit once a month in a meeting with that person, and by nature they interact better with solid, predictable people.

But a huge part of it is lack of understanding of how much software development differs from other occupations. Software is an art, not a form of engineering: much more like music than like building bridges. We have long known this, and yet the realization has never completely sunk in.

Music is an apt simile in a special way: namely that there was a point in time where musicians were treated like mere servants. While painters, sculptors, and architects enjoyed patronage and fame, musicians were largely obscure, confined to their local environment, and treated much like kitchen staff. They would exchange information among themselves, but nobody in the wider world cared much about musicians.

This changed in the 18th century, as the middle class started growing and growing powerful. While until then the courts had ruled music, there was now a growing market for sheets and instruction from the nouvau riche. The decoupling of producers (of music) and employers had the immense effect of decoupling revenue and popularity at the same time. The star musician was born.

No other musician is a better example of the transitional figure than Mozart. In his youth, most of his and his father’s time was spent chasing after employment at one court or another. The Archbishop of Salzburg, in whose employ both father and son were, was incensed at the arrogance of his employee, who would take leave of absence whenever he felt like it. The eventual break was celebrated with a kick in the rear (an actual one, administered by the Archbishop’s stewart, a count himself).

Mozart didn’t fare well in this new world. He did have huge income sources and a set of powerful patrons, but apparently he was good at spending money, and his music eventually lost favor. Add to that the French Revolution, and you find a composer that didn’t finish his life with the fanfare he probably deserved. His successors, though, especially Beethoven, made the new world order permanent. After Mozart, the idea of a servant composer seemed quaint.

I believe we are rapidly approaching a similar transition in software. We have long known that what is called “star developers” are several orders of magnitude more productive than average ones. Plodding developers, on the other end of the scale, cause more harm than they do good, and any company would be better off terminating their employment from a pure productivity perspective.

The main reason software is so different from, say, civil engineering is that it requires rapid adaptability. You need a core set of ideas, processes, and tricks that you are able to adapt and apply to whatever new environment you are thrown in. You need to be able to challenge your own preconceptions and adopt new ways in a critical fashion. You need to know the best tool for the task, and you need to be able to apply it.

Any for of art requires three core abilities: invention, skill, and knowledge. You have to know your tools and options; you have to have the mechanical ability to perform the task, and most importantly you need to have the creativity to come up with something better than what is there.

In software development, the situation is the same. Certainly, the mechanical skills are limited to the ability to type rapidly and accurately. But knowledge of the trade is of paramount importance, and creativity is incredibly important.

The problem, nowadays, is that those handing out the money do not understand the art, nor are they willing to reward it. Until it is understood that a CTO is more important to an Internet startup than a CEO (and the share allocations are handled accordingly), the only way to be successful in an Internet startup is to create one.