=============================================

[From The Art of ‘Ware (Version 2.0) by Bruce F. Webster (forthcoming), Chapter 2, “Supporting Development”]

In key areas of technology development, talent is scarce and salaries are high, limiting resources for other employees.

As new technologies become hot markets, the number of skilled developers is small, and they command high salaries. For example, in the software industry this has been true at various times for 8086 assembly, Windows, C++, OLE 2.0, object-oriented development, Java, .NET, Python, and subsequent technologies. Each area fills in with time as sustained demand and high salaries draw more engineers into it. But, curiously enough, the absolute number of excellent developers in a given area of technology remains pretty much the same.

When I was teaching computer science at Brigham Young University in 1985-87, the number of students enrolled as computer science majors had increased dramatically — by a factor of five or so — from when I had been a student there a decade earlier. One of the professors, who had been around since the early 70’s, observed to me that the number of really good students in the department was still pretty much the same; the five-fold growth of enrollment hadn’t brought a five fold, or even a two-fold, increase in excellent CS majors. Why? Because those students with interest, aptitude, and native talent has been signing up all along; the surge in enrollment had come from students who saw computers as a way to get a great paying job, much as my friends during my undergraduate days had signed up for pre-law or pre-med.

When a new area of technology opens up, it quickly draws to it those developers with the interest, talents and desire to become really good in it. More excellent developers do come along with time, but as they do, some of the current ones start moving on to new areas, so the absolute number stays more or less constant.

Because of this, I’d like to propose a minor addition to the vast assortment of laws and rules governing technology and engineering:

• Webster’s Constant: the number of excellent developers in a new area of technology quickly reaches a constant value, which is sustained through the period during which the technology is vital.

This may seem a bit silly or fatuous, but it’s actually critical to understand if you need to build a development team that will be working with key technologies. It’s going to be tough finding really good people, and you’ll find yourself running into the same names over and over again. The trick is getting them to come to work for you.

==========================

Compare suntzu pingfa (Chapter 2: “Doing Battle”):

A nation can be impoverished by the army when it has to supply the army at great distances. (Sonshi translation)

Those who handle product development skillfully don’t build engineering teams twice, nor raise capital three times.

Building product development teams twice means having to replace the original engineers with new ones in the order to complete the product. There can be any number of reasons for having to do this: the engineers get burned out; the engineers get disgusted with upper management and leave; the engineers lose faith in the company and its directions, particularly if they view the product as being “hijacked” and taken in a different direction by latecomers to the company, or the engineers are replaced and/or fired, either because of insubordination or because they weren’t the right ones for the job in the first place.

Raising capital three times before product release indicates that development and launch have taken too long. (Believe me, I know.) The first round is usually essential to get the company off the ground. The second round may be necessary because of changes in product direction or the all-too-common delays in production development. But you’re in trouble if you raise a third round of capital for anything but product launch, and possibly even then. Not only does that mean that you’re late in shipping, it also means that you’re surrendering equity — thus reducing equity incentives for existing employees — and that you do not have sufficient cash reserves to keep the company going once the product does finally ship.

Focus on creating your own development resources, and recruit from your competition. This way, you can be sufficient in both tools and personnel.

Invest in development resources, that is, the tools needed to actually create the product. All the money you can think of possibly spending on those resources probably won’t equal what you’ll lose each and every month if your product is late. Likewise, be willing to devote people and resources to creating custom in-house development resources. It’s easy to chose not to do this, because you can often “get by” without such tools and you may be concerned about the return on investment. But the right tools can make significant difference in product quality and time to completion.

Some of the best people to do your product development are at other companies and are probably at your competitors. Recruit aggressively and hire the best, strengthening yourself and weakening your competition.

=====================

Compare suntzu pingfa (Chapter 2: “Doing Battle”):

Those skilled in doing battle do not raise troops twice, or transport provisions three times. ?

Then the army will be sufficient in both equipment and provisions. (Sonshi translation)