[cletus]

Honestly this is just a PR piece for DeveloperAuction. "How to win the bloodthirsty battle for tech talent"? Why come use our site of course!

This strip from 1995 sums it up pretty well:

http://dilbert.com/strips/comic/1995-05-22/

I see a number of problems (on both sides):

1. There is no shortage of talent. There might well be a shortage of talent for the price you're willing to pay but that's hardly the same thing;

2. What constitutes "talent" seems to be largely based on social proof. 25 year old Stanford graduate? Offers galore I'm sure compared to, say, the 45 year old University of Iowa graduate. Hell, I get an awful lot of cold calls based simply on listing "Google" on my online profiles (as my employer).

Social proof can be a useful indicator. The problem is that groups tend to self-select down to nothing this way where you end up with a tiny fraction of the group being over-subscribed and the majority struggling;

3. People like to employ people like themselves. So find a company full of MIT graduates and they're likely to hire... more MIT graduates. This isn't just a question of social connections or geographical area either (IMHO);

4. If you pick a high-demand high-cost area like the Bay Area you're obviously going to have a harder time finding and retaining talent and it'll be more expensive;

5. Larger companies tend to treat talent as interchangeable where the only units are the number of warm bodies, perhaps stratified into "junior", "midrange" and "senior" whereas we all know there can be a 10x or greater difference between two engineers in terms of productivity, hence the more productive talent is harder to attract and retain.

[KaeseEs]

The 10x figure is pseudoknowledge - something we "all know" that isn't actually true (or that at least we have no good reason to believe to be true), cf. http://vimeo.com/9270320 . In brief: the study that it is ultimately derived from had a tiny sample size (on the order of 30, to wit), and for several reasons the experimental design would have invalidated its results outside the very narrow original focus of the study (which was the relative productivity of batch processing vs. interactive processing on the computing systems of 1968) even if the sample size had been larger.

[JoeAltmaier]

The 10X figure is true. My study is 30 years of working with a wide variety of people in startup environments. Adding some people makes a project later. Others can manage a whole project alone (if they are LEFT alone).

[lifeisstillgood]

I would like to think there is a talent shortage.

I pitch that we are in a similar position to Europe post Gutenberg - where they went from 2% literacy rates to 20% in a gut wrenching century. We are going to go from a lot less than 2% of people who are source code literate to that magic 20% in less - in short we are millions of talented people short.

In our generation, the observation talent shortage only occurs at certain ,price points is right - but when you want all companies to have the same code in their DNA as perhaps google or MS, we are so embarrassingly short of people it's a public policy issue not a Market lead one.

But I still laugh at the filbert cartoon - it's just I acknowledge that is true if your definition of talented developer is a software engineer who will write code primarily for a living and perform a sort of back office function. What about the CEO who codes? The deputy CFO?