Let’s assert that there are two kinds of jobs you need to fill:

The first kind of job is a cog job. A job where you need someone to perform a measurable task and to follow instructions. This can range from stuffing envelopes to performing blood tests. It’s a profitable task if the person is productive, and you need to find a reliable, skilled person to do what you need.

The second kind of job requires insight and creativity. This job relies on someone doing something you could never imagine in advance, producing outcomes better than you had hoped for. This might include a sales job, or someone rearranging the factory floor to increase productivity. It could also include a skilled craftsperson or even a particularly skilled receptionist.

If you’re hiring for the first kind of job, exactly why are you sitting a nervous candidate down in your office and asking her to put on some sort of demonstration in her ability to interact with strangers under pressure? Why do you care what his suit looks like or whether or not he can look you in the eye?

If you’re hiring for the second kind of job, the question becomes even more interesting. Would you marry someone based on a one hour interview in a singles bar? And how does repeating the forced awkwardness of an interview across your entire team help you choose which people are going to do the extraordinary work you’re banking on?

I’ve been to thousands of job interviews (thankfully as an interviewer mostly) and I have come to the conclusion that the entire effort is a waste of time.

At least half the interview finds the interviewer giving an unplanned and not very good overview of what the applicant should expect from this job. Unlike most of the marketing communications the organization does, this spiel is unvetted, unnatural and unmeasured. No one has ever sat down and said, “when we say X, is it likely the applicant understands what we mean? Are we putting our best foot forward? Does it make it more likely that the right people will want to work here, for the right reasons?” [tell the truth, do you test your job interview spiel the same way you test your web results or even your direct mail?]

The other half is dedicated to figuring out whether the applicant is good at job interviews or not.

I should have learned this lesson in 1981, when my partner and I (and three of our managers) hired Susan, who was perhaps the best interviewer I have ever met. And one of the worst employees we ever hired. Too bad we didn’t have a division that sold interviews.

Let me be clear about what I’m recommending: the next time someone asks you to “sit in” on an interview, just say no. Don’t do it. Don’t waste your time or theirs.

So, what should you do instead?

Glad you asked!

First, none of this will work if you’re not offering a great job at a great company for fair pay. These techniques will not succeed if you are the employer of last resort. Assuming that’s not the case, how about his:

Every applicant gets a guided tour of your story. Maybe from a website or lens or DVD. Maybe from one person in your organization who is really good at this. It might mean a plant tour or watching an interview with the CEO. It might involve spending an hour sitting in one of your stores or following one of your doctors around on her rounds. But it’s a measurable event, something you can evaluate after the process is over. If you’re hiring more than a few people a week, clearly it’s worth having a full-time person to do this task and do it well.

There are no one-on-one-sit-in-my-office-and-let’s-talk interviews. Boom, you just saved 7 hours per interview. Instead, spend those seven hours actually doing the work. Put the person on a team and have a brainstorming session, or design a widget or make some espressos together. If you want to hire a copywriter, do some copywriting. Send back some edits and see how they’re received.

If the person is really great, hire them. For a weekend. Pay them to spend another 20 hours pushing their way through something. Get them involved with the people they’ll actually be working with and find out how it goes. Not just the outcomes, but the process. Does their behavior and insight change the game for the better? If they want to be in sales, go on a sales call with them. Not a trial run, but a real one. If they want to be a rabbi, have them give a sermon or visit a hospital.

Yes, people change after you hire them. They always do. But do they change more after an unrealistic office interview or after you’ve actually watched them get in the cage and tame a lion?