Rather Test or Guess?

“Make me a deal on a split run.” Of all the negotiating ploys we as marketers might consider, this simple sentence has more success-seeds than any of the fustian and fury we could force out of our bargaining-parleying fingertips. And a “Yes” answer from an understanding medium, which costs zilch, has to result in information far more profitable than even our top-of-the-line brainpower can match.

TM0810_searchglobe copy“Make me a deal on a split run.”

Of all the negotiating ploys we as marketers might consider, this simple sentence has more success-seeds than any of the fustian and fury we could force out of our bargaining-parleying fingertips. And a “Yes” answer from an understanding medium, which costs zilch, has to result in information far more profitable than even our top-of-the-line brainpower can match.

One assumption we certainly have enough professional knowledge to lean on: the circulation of the medium has at least a tenuous match with a logical buyer. Our prospects won’t think we’re approaching from the planet Mars.

For print media, a split run is easier to mount today than it ever has been since, some hundreds of years ago, we as marketers invaded the nooks and crannies of publishing. For direct mail, it’s a bonanza whose luster dimmed when direct radio and then direct television mussed up the turf. For online, it’s too natural and obvious to be regarded as an innovation.

The overriding interpretation of what we’re discussing is a single word: test.

If the notion of testing a direct appeal is foreign to you, call me or any of about fifty thousand other self-proclaimed marketing experts, and we’ll be glad to take advantage of your naiveté.

Or, if you’d rather, make one decision that has to be profitable: what to test.

The most common test element is price. What price represents the best addition to the bottom line? $19.99 may bring more orders, but $24.99 is more profitable. And in today’s wild marketplace, where 99 cents has almost universally replaced the venerable 95 cents, $24.99 just might bring in more responses than $19.99. What if we glamorize the offer? $29.99 versus $19.99? We can test to give us an answer.

(Sample example: a recent three-way test for a collectible priced the item at $15.49, $15.99, and $17.99. Which brought the highest total number of responses, not just dollars? Right. $17.99. I suspect because the product has a tie to tradition, $17.95 would have left $17.99 in the shade, but the testing impulse didn’t extend that far. Maybe next time.)

And easy? What test could be easier? Just be sure that each addressee gets just one distinct offer and the response code differs for each price.

“Seat of the pants” guesswork is old-fashioned and amateurish, and depending on the deal you can make with media or a lettershop, not an optimal investment in marketing.

Hmmm. Here’s a unisex jacket. Here’s a tablet computer. Here’s a DVD whose content dwarfs any approach to the business problem its content solves. Here’s an extraordinary assortment of dessert-goodies.

A true split is just one split-test: When an offer appears on our monitor, we can’t tell if it’s unique or part of a split run … that is, if the code doesn’t betray the technique.

What does that mean? Well, suppose you get an online offer from “Firearms.” Does that, emotionally and in your mind factually, differ from “Guns” or for that matter the singular, “Firearm”? What if the sender had split the subject line, sending to one group “Look out. This gun fires in both directions” and to a parallel group “Gigantic 75% discount, today only.” Even from this example, any of us can predict that response will be skewed by the difference in appeals. What we have is a message test, even though only the subject lines may differ.

Testing for B-to-B Marketers: How Hard is It?

B-to-B marketers are often guilty of laziness when it comes to testing their communications, whether it’s testing the copy approach, the layout, the offer or the target audience. Well, to call it laziness may not be entirely fair. It’s a fact that the typical B-to-B campaign targets universes that are too small to support a split test. If you’re selling specialized machine tools, you’re lucky if you have 10,000 potential customers worldwide.

B-to-B marketers are often guilty of laziness when it comes to testing their communications, whether it’s testing the copy approach, the layout, the offer or the target audience. Well, to call it laziness may not be entirely fair. It’s a fact that the typical B-to-B campaign targets universes that are too small to support a split test. If you’re selling specialized machine tools, you’re lucky if you have 10,000 potential customers worldwide.

I work with a company that offers employee benefits programs, and markets to HR professionals. We are planning a campaign to take the service into the Boston area, targeting firms with more than 100 employees, which number about 6,000 sites. At two HR contacts per site, using direct mail, we would have a mail plan of 12,000. With an estimated response rate of 1 percent, we’re looking at only 120 inquiries-clearly not enough to conduct a test of the two good offer ideas we are kicking around. Which is a shame, because we really have no idea which motivational offer is going to work better with this audience.

But in the digital world, B-to-B marketers have a lot more options for testing. Split tests are easy to set up, and applicable to any communications vehicle that drives a response-whether it be an email, a landing page, a banner ad, Adwords copy, anything, using free tools like Google Website Optimizer or scores of other SaaS or enterprise software tools.

Plus, there are abundant resources out there now to guide and inspire business marketers. Have a look at Which Test Won, a weekly comparison of two B-to-B live test versions-usually landing pages-where visitors are invited to go with their guts, and pick a winner. Then, you can view the actual winner and participate in a lively discussion of possible reasons why. This brilliant site was the brain child of Anne Holland, the founder of Marketing Sherpa.

So my client would like to conduct an offer test through digital channels, and we are exploring various options. It’s still not easy with a small prospect universe in a limited geography. There are not enough targeted banner media available to reach HR professionals in the Boston-only area. Email to entirely cold prospects is too spammy to generate leads at a reasonable cost-and still doesn’t solve the universe size problem that we face with direct mail. We considered Google AdWords with location targeting, but it’s going to be hard to sell the offer properly within the AdWords copy limits. Not to mention questions about how long it would take to get enough clicks to call the results. So our search continues, and we’d welcome ideas from Target Marketing readers on this one.

A version of this article appeared at Biznology, the digital marketing blog.