If you’ve looked for a job recently you know the awful statistics: fewer than three out of every ten jobs are filled through normal channels.

Let me translate this for you: trying to get a job by sending your resume to Human Resources in response to an employment ad is a sucker bet.

The system is broken … badly broken … and the numbers prove it. What’s truly pitiful is that hiring managers don’t like the current state of affairs any more than job seekers do. If they did, the numbers would be very different.

Don’t believe me? Keep an eye out for “Ask the Headhunter: reinventing the interview to win the job” by Nick Corcodilos, which should be hitting the bookshelves this August. Nick has been in the headhunting game a long time, and has succeeded by ignoring most of the nonsense spouted by what he calls “the employment industry”. As Nick points out, “You will encounter many people who are not really the person who will hire you – they are the go-betweens who want you to hunt for a job in a way that’s convenient for them.”

Actually, he’s talking to both the applicant and the hiring manager, because when you’re hiring you’ll also deal with go-betweens.

That’s exactly what you want from HR, whether you’re looking or hiring: To connect the applicants most likely to succeed with the hiring managers who need them. Far too often, HR screens out the very people most likely to succeed instead: people who are stretching, who want a new challenge, who haven’t done the job you’re posting but who will do whatever it takes to succeed at it.

What’s the problem? In most companies, HR has an unstated mission: keep the company out of court. It does so in any number of ways: ensuring compliance with various employment laws; creating personnel handbooks so everyone “knows the rules”; helping managers define position requirements in terms of “objective” evaluation criteria; screening resumes to ensure hiring is done by strict skill-to-task matching … (which is now an automated process, give me strength!).

Keeping the company out of court is a Good Thing (GT, to use the acronym). Of course, people will sue you anyway, and in the meantime you’ve hired and promoted a lot of the wrong people, damaging your company’s ability to compete.

(Now before you flame me, let me draw a clear distinction between individual human resources professionals and the HR industry. I have quite a few friends who work in HR and as a whole they’re goodhearted people who seriously want to help both their employer and their coworkers succeed. Few are given a chance: their industry conspires to prevent it.)

Years ago a friend of mine, new to management, asked the most important consideration when hiring. “Hire a person, not a resume,” I told him. “The skills you’re looking for today won’t be the ones you’ll need next year, so find people with the right aptitude and a habit of succeeding. They’ll acquire whatever skills they need to succeed. Even better, they’ll do the jobs that need doing, not just the ones you think are important.”

I still think that was good advice. Here’s some more: when writing a job description be specific when it comes to attitude and tangible results, and as general as you can when defining skills. If you’re hiring a database administrator, for example, you don’t want someone who will turn into the “data police” and do want someone who thinks of the job as a way to make programmers more effective. Do you really care that her ten years of experience are in Sybase and Oracle while you use Informix?

Turn it around: if you’re a database administrator who knows Sybase and Oracle, do you avoid positions that will cause you to use Informix?

Nick Corcodilos will tell you more: that both applicant and hiring manager need to conduct interviews that are about doing the job. The applicant should do the job in the interview. The hiring manager should ask the applicant to do the job in the interview.

Because, in the end, you want to hire someone who can do the job, not someone who can do the interview.

Right?

Back in my electric fish days I helped my French friend “Bullet” Micheloud develop a computer simulation.

Bullet had laboriously collected sex ratio data on two species of fig wasps (if you like Fig Newtons you don’t want to know – trust me). We ran the simulation on a Texas Instruments programmable calculator because I was a weenie, but not enough of a weenie to use a Hewlett Packard. We based our simulation on the theory of sex ratio selection, and predicted his results pretty well.

Back then we knew the difference between mainframes, minicomputers, and microcomputers: Mainframes had 32 bits, minis had 16 bits, and micros had 8. It was simple, clear, and easy to keep track of.

Now it’s all a matter of opinion, and it’s my opinion that the mainframe is dead, despite the thirty-seven articles you’ve just read declaring “The Big Iron is Back”. It’s IBM that’s back, and you can learn from its strategy … after I put another nail in the mainframe’s coffin.

The mainframe really is dead, and IBM just killed it. It still sells boxes it calls mainframes, but that’s just marketing.

Have you seen these puppies? They’re the size of a big file server. They’re built around a bunch of microprocessors. They include the whole AIX API, according to IBM – in other words, UNIX.

IBM can call this a mainframe if it wants, it can tell me it now runs AIX on its mainframes, and it can repeats its mantra that “the mainframe is just the biggest server on the network”.

Heck, IBM can explain that teenagers from Arcturus get bored on Saturday nights, fly to Earth, make big circles in cornfields, and then laugh hysterically at the stupid Terrans who come to gawk.

It can tell me this all it wants, but that won’t make me believe it.

IBM could have announced that it had ported MVS to its midrange computers instead. It’s all in how you look at it, of course, and if it makes Big Blue happy to call its new superserver a teeny weeny mainframe, well, so be it.

Regardless, we’re all going to wake up one day soon and discover something remarkable. IBM, while we weren’t paying attention but in plain sight, will have transformed its entire product line into a single platform, built around PowerPC technology. All of its boxes will be able to run MVS, AIX, and OS/400, and Windows NT as well.

Who knows? Rhapsody may run on it, too, now that Apple has survived Larry Ellison’s silly takeover idea. (I never could understand why someone who hates the idea of stand-alone PCs as much as Ellison wanted a company that caters to the most rabidly independent PC users on the planet.)

This forecast doesn’t bode well for OS/2, of course, but very little has bade well for OS/2 since IBM took it over from Microsoft.

Now about that career lesson:

Think of the technology marketplace as an employer, and IBM as a corporate executive who took some highly public mis-steps and found his career in the doldrums. It happens in every large company, and it can happen to you.

If you don’t want to find other employment and you do want to resuscitate your career, you can take some lessons from IBM right now. IBM first did everything wrong, announcing semi-annual reorganizations, pricing changes, and a variety of incomprehensible strategies – it was in denial. It has since found the right answer: Great engineering and solid execution. It has started to stick to its knitting so it will have real successes to point to, while the nonsense fades into memory and its rival overplays its hand.

If your career has taken a wrong turn, follow IBM’s current strategy: Stick to your own knitting and amass some clear successes. Ask for a project or problem area to clean up, drop out of sight for awhile, and do a lot of stuff right. Eventually, your own nonsense will fade into memory, and your own rivals will overplay their hands.

There are worse ways to succeed.