marketing math

For a Conversion Scientist, life is full of A/B tests. You find them everywhere. We are naturally curious creatures, always wondering why people behave the way they do. The answers are rarely obvious.

For example, I have been monitoring a sort of A/B split test happening at a local cupcake and coffee shop called Cupprimo. For some unknown reason, the proprietors have consistently provided both quilted and unquilted toilet paper in their men’s room. As you might expect, the quilted TP roll is consistently smaller than it’s unquilted competitor.

But, as a Conversion Scientist, I have to ask: is it because the position of the quilted roll is more convenient to right-handed bomb-droppers? Is it because the luxurious quilted TP is thicker, and that the roll only appears to shrink faster? Perhaps it is a test error? Are the hapless managers not replacing the rolls at the same time? Here’s the real kicker: are these results consistent with the ladies’ results?

I know that, like me, you’re just dying to find out what the test really tells us. But right now, I’m going to spend the next few paragraphs trying to tie this example to my presentations at the Door64 Tech Fair and the Extraordinary Interviews Seminar for Technology Professionals. You want to go, you just don’t know why yet.

The Job Search and Conversion

While the toilet paper question is important, I’ve had to focus my attention on the problem of job search for the past few months. As in our toilet paper split test, I am dealing with incomplete information in the job search. Elusive are the conclusions we would draw from our job search data.

The ultimate conversion is a job offer. However, it turns out that the resume-to-job-offer conversion rate is fraught with trouble, the most significant being that once we get one or two conversions, we stop caring. We take a job and stop searching.

What we need for the job search is predictive metrics. For most job seekers, the most important predictive metric has been number of resumes sent. But this approach is no different from that taken by any number of companies who’ve spent like drunken sailors buying clicks for search terms that have little to do with their business, hoping that someone might actually become a lead or buy something–anything. This is not conversion.

A more important measure is number of interviews. For the sake of conversion, we measure it as the number of interviews booked divided by resumes sent. I call this the resume-to-book rate. The drunken sailor set will say, “So the more resumes we send, the more interviews we will get, right?” Well, since the resume-to-book rate is a percentage, throwing resumes against the proverbial wall will actually decrease your conversion rate (unless you send each with a substantial sum of money, which is also true in pay-per-click search).

The key is to send fewer more targeted resumes.

When you do this three things increase your resume-to-book rate:

• You decrease the number of resumes sent, which will mathematically increase resume-to-book rates.
• You increase the likelihood of getting more interviews by spending more time on each application and cover letter.
• You target only those jobs for which you are somewhat qualified.

In other words, your doing the exact opposite of what most search marketers do.

Meaningful Job Search Conversion Rates

It turns out that, like our quilted toilet paper test, our resume-to-book rate has some built-in issues. Since 80% of all jobs offers are garnered from a personal contact, sending resumes is actually the least effective way to apply for a job. The number of postings available via job boards is too small and there’s too much competition. In short, it doesn’t scale. The Conversion Scientist ranks #1 on all search engines for “conversion scientists with lab coats.” However, the number of searches is small, so it really doesn’t help. Job postings have the same problem.

Since personal contact is the way to get interviews, then our predictive conversions must be tied to the number of people who are actively helping us look for work. For most of us, that is our parents (motivated to get us to move out of the house again) and that one new recruiter who hasn’t yet learned a tactful way to tell us they don’t really have anything for us.

As it turns out, it’s our old friend “opt-in rate” that is the most powerful predictor of our job search. Most of us have low opt-in rates. ‘We’ve Twittered, “Got laid off. Let me know if you hear of anything.” That doesn’t invite an opt-in. A few of us have sent emails to our professional contacts detailing our qualifications. Then we end by saying “Let me know if you hear of anything.” Not an opt-in either. In short, we don’t give our personal networks anything to opt into, and our opt-in rates would be zero if not for our parents and that one green recruiter.

Why is opt-in important? Because the higher your opt-in rates, the larger your personal job search network gets. The larger your personal job search network, the more unpublicized jobs you’ll be privy to. The people who know how to engage their personal network have the following advantages:

• They uncover jobs that aren’t posted
• They get inside information on the jobs they apply for
• They find champions to walk their resumes into the hiring manager
• They circumvent the trolls in HR whose main job is to protect the hiring manager from tidal waves of resumes