I wonder at the ambitiousness of your recruiting vs. the salaries you pay. Can you reasonably expect to get a world class programmer for 100k and stock options?
According to your criteria, you're thinking more of 'potentially world class', which I guess is more in line with your price range. The world class guys work for Google.
I think if you plot track record vs. productivity you'll end up with a (largely) normally distributed curve. If you plot track record vs. cost, you'll get a curve that rises exponentially. Eventually, talented people get paid exorbitant salaries based entirely on their previous accomplishments, and these are exactly the types of people you don't want to hire (because their cost isn't justified by their productivity). It's logical, then, to form a world-class team of first employees largely composed of diamonds in the rough, or 'potentially world class' as you put it.
> these are exactly the types of people you don't want to hire
Depends on your budget and appetite for risk. If you have the budget and want to reduce your risk, IMHO there's no better indicator of future performance than past performance.
> If you plot track record vs. cost, you'll get a curve that rises exponentially.
I've got a pretty good track record. How do I get to work in your world? :-)
IME, for regular employees, the difference in pay between a respectable senior developer, technical lead or project architect and a junior developer straight out of university is perhaps a factor of 3-4x. Unless you go into management and rise to a senior level, it pretty much plateaus after that in most places.
On the other hand, the difference in productivity between the same types of people can be a lot more than 3-4x. By the time you factor in the improved communications and general reduction in overheads from having a smaller team of individually more productive people, top people are easily worth a disproportionate amount, if only you can find them and convince them to work for your company instead of someone else's (including their own).
Comments
I wonder at the ambitiousness of your recruiting vs. the salaries you pay. Can you reasonably expect to get a world class programmer for 100k and stock options?
According to your criteria, you're thinking more of 'potentially world class', which I guess is more in line with your price range. The world class guys work for Google.
I think if you plot track record vs. productivity you'll end up with a (largely) normally distributed curve. If you plot track record vs. cost, you'll get a curve that rises exponentially. Eventually, talented people get paid exorbitant salaries based entirely on their previous accomplishments, and these are exactly the types of people you don't want to hire (because their cost isn't justified by their productivity). It's logical, then, to form a world-class team of first employees largely composed of diamonds in the rough, or 'potentially world class' as you put it.
> these are exactly the types of people you don't want to hire
Depends on your budget and appetite for risk. If you have the budget and want to reduce your risk, IMHO there's no better indicator of future performance than past performance.
> If you plot track record vs. cost, you'll get a curve that rises exponentially.
I've got a pretty good track record. How do I get to work in your world? :-)
IME, for regular employees, the difference in pay between a respectable senior developer, technical lead or project architect and a junior developer straight out of university is perhaps a factor of 3-4x. Unless you go into management and rise to a senior level, it pretty much plateaus after that in most places.
On the other hand, the difference in productivity between the same types of people can be a lot more than 3-4x. By the time you factor in the improved communications and general reduction in overheads from having a smaller team of individually more productive people, top people are easily worth a disproportionate amount, if only you can find them and convince them to work for your company instead of someone else's (including their own).
if you plot track record vs. productivity you'll end up with a (largely) normally distributed curve
Are you saying people with great track records are likely to be very unproductive? That seems unlikely ...