Engineer Compensation
Lots of people have asked me about this, sorry for the delay, figured I'd just write a post instead of respond individually.
Let's clear away some of the misconceptions: The WRONG way is to lowball them in compensation and try to sell them SOLELY on "the vision". While most talented engineers do have a passion for creation, they are certainly not the engi-nerd stereotypes you see on TV, and most of them would just feel insulted.
At the same time, don't blow their contribution out of proportion. Non-technical founders either go too low (and can't understand why engineers turn them down), or overpay.
Engineers are excellent at assessing opportunity cost. The more qualified they are for your space, the better they are at assessing the likelihood of your success and figuring out how much the options you're giving them are worth.
Another terrible argument: I'm giving you stock so the long-term upside is enough to cover the extremely low pay.
….Actually, they can easily come up with a dollar value of the shares you are offering, vested over 4 years, and compare it to the lost wage/bonus per year multiplied by 4. Most of the time, the value of your options are WAY lower.
That's not to say you have to compete dollar for dollar, that's impossible; you won't beat the Google, Amazon, or Facebooks of the world. But you have to understand the approximate value and structure your recruiting pitch accordingly.
The right level of compensation should be based on a combination of what you need them to do now, and what you want them to do in the future. My approach is…
Salary - calculated based on their current level and the role I want them to play. I go market rate or very close to it.
Bonus - calculated as a percentage of the salary I offer. They'd get a bonus anywhere else they worked.
Stock - calculated as basis points on their seniority and the role I want them to play. They should have a stake.
Be generous, and be equitable, and you'll be able to entice those stellar engineers. (Finding them however, is another problem… :P)