logoalt Hacker News

msteffen • today at 5:51 PM • 1 reply • view on HN

I agree that support is hard.

Since you seem to have experience with this, the principle that I've seen work "best" in my own limited experience (and I wonder how it matches with your experience) is to keep the barriers between support and product development as low as possible:

1. keep the company as small as possible in general (which lowers all communication barriers everywhere, and helps with cash flow—fewer salaries, smaller office, low overhead)

2. Early on, have the same engineers doing product development and support. Later, have heavy, regular communication between the teams (including managers and senior ICs on both sides. Maybe rotating the managers would work? I don't know. Once you have more than one level, I don't know anymore, but maybe do the same thing with directors and senior managers?). This hurts growth, because you're understaffed to begin with, and you're impeding the staff you have by giving them tons of unplanned support work. But IME it means that product development is happening at the right level of abstraction.

If the product you're building requires big company support (or training or product development), and the revenue it brings in can only support a small company, then you should change what you're building. It's very hard to change what you're building if the company is very big.

So if, for example, support is boiling because your product is flaky, and engineering is too small and too busy to fix all the problems at a reasonable pace, you probably have a design problem(s) and expanding will make it more entrenched. Instead, you should simplify the product. Then expand (with an understanding of the real costs of different parts of your vision), and repeat.


Replies

justin66 • today at 6:50 PM

Developing cold fusion is hard. Support is a solved problem that companies fuck up because they don't want to spend the money to do it properly.