Brent Hodgson, Copywriter

If you’ve been reading my blog for the last few months, you know that we split test a long form salesletter-style job ad recently.

You also know that I promised to post up the results of the split test - but I’ve been really slack and haven’t got around to it yet.

Planned Methodology for Testing the Salesletter

We took one standard style job ad - full of all the usual three-letter-acronyms and corporate chest-beating that you would find in any job ad.

And then we wrote a long form sales letter for the job which was advertised (an analyst programmer position).

The salesletter version spoke about the benefits of working at Alliance Software, included testimonials from current employees talking about why they enjoy working here - and so on.

It was great - I could feel it - the salesletter was going to stick out like a sore thumb!

It couldn’t have been more different to all the rest of the ads out there.

We put the salesletter and the standard job ad up on one of Australia’s biggest job-ad sites, and started measuring the source of job applications which were sent in.

But the Long Form Salesletter didn’t perform as planned!

Unfortunately, we noticed at the beginning of the split test that we were receiving far more applicants via the standard job ad than the long form salesletter version.

We found out later that one of the ads (the standard job ad) was displayed in a much more prominent area of the job ad site to the other ad (the long form salesletter).

Rats!

This meant we weren’t able to complete a true split test

However… When we reviewed the data, it showed something very interesting.

The Salesletter attracted higher quality job applicants

Even though the standard job application received significantly more applications, the long form salesletter seemed to attract a better quality of applicant.

In fact, half of the applicants who made it to the interview stage applied via the long form salesletter - despite it only receiving a small fraction of the total number of applicants.

I bet you’re thinking:

“Yeah, but you could have been influenced to interview more people from the long form salesletter pile.”

Good point - knowing the source of the application (and subconsciously wanting the salesletter to win the split test) could have skewed the results… If the source of the application was known while they were being reviewed.

But each application was reviewed on its own merits - and there was no way to tell whether the application came in via the salesletter.

We only reviewed which applicants had applied via the salesletter after the interview process was over.

Unfortunately, this wasn’t a true split test.. but I’m looking forward to giving the long form salesletter another shot to see whether it can create an avalanche of high calibre job applicants next time.

Popularity: 5%