Resolve to Test in 2005

Today I’m offering you two holiday gifts. First is a rundown of some thought-provoking marketing tests. You can benefit from the tests without having to conduct (and pay for) them. Second is a free statistical calculator to help you analyze the tests I hope you conduct in the coming year.

What Is a Picture Really Worth?

Craig Garber is a direct-response copywriter and direct-marketing consultant. He conducted a simple test on his Web site. The test boosted his already strong sign-up rate by another 5.35 percent.

The site’s entrance is a “name squeeze” page, designed to get visitors to sign up for Garber’s free “Tip Of The Week.” People usually come to the site because they either read something he wrote and liked it or were referred by a friend or colleague.

In a side-by-side comparison, when Garber added his picture to the front page, his sign-up response went from 45.65 percent to an even 51 percent.

“When your prospects can ‘see’ you’re a real person, they make a connection with you that increases their trust and belief in who you are and what you’re offering,” said Garber. “Especially online, where most people don’t even give you contact information, let alone a personal glimpse into [their lives].”

As a copywriter, Garber’s a firm believer in testing. He almost always has several tests going at the same time.

Should You Test Something New?

A major online banking concern begrudgingly agreed to conduct a test. It had a working marketing process that brought in new account applicants for about $35 per acquisition. The company was talked into testing a few new potential lists and sources for its email acquisition campaigns.

Much to its surprise, some of the tested lists had costs-per-acquisition of less than $20. This is a huge improvement. Sometimes you don’t know until you test.

Creative Doesn’t Always Create

Test results aren’t always what you expect to see.

One company with a very expensive big-ticket product conducted a customer acquisition test using two different creative approaches. One approach was a standard mail piece that cost about $1 per piece to create and mail. The other was a much fancier piece of creative and was sent via FedEx. It cost about $12 a piece. Test results showed the additional expense didn’t result in any response rate increase.

Is that surprising to you?

No, We Don’t Need the $5 Million

My all-time favorite test result came from a direct mail campaign’s data analysis. The test was conducted by a major long distance telecommunications company. Now, I’ve worked on projects for quite a few telecommunications companies, but this result amazed me more than any other I’ve seen.

The company tested two versions of a mail piece, with only one difference. The fine print of one version mentioned the company would cover the local phone company’s costs for switching to the long distance company. The other version didn’t mention the fee.

Results showed clearly that covering the fee did nothing to increase response rate. The implication was the company could save about $5 million in the coming year by taking the switching fee out of the current offer. Astoundingly, the campaign managers did not make the change that would save the company money. Why? The money didn’t come out of their budget.

Ready to Test? Get This Tool

All marketers, especially online marketers, can and should conduct tests. Simple tests are easy to design, conduct, and analyze. You can determine whether the results are statistically significant with basic statistical analysis. My second gift to you, a free statistical significance calculator tool (scroll down a bit on the page), can help you with this analysis.

Soon we’ll be making resolutions for next year. I hope your business resolutions will include a commitment to conduct more tests to improve your business.

Related reading

Big Data & Travel
Flat design modern vector illustration concept of website analytics search information.