Consider randomized, real-time testing

I think all marketers have had a similar experience: You're meeting with a client/superior/colleague discussing an upcoming campaign. Your client is absolutely sure that one offer/creative will outperform another. You, on the other hand, just KNOW that the alternative offer/creative is the winning package. You argue about the benefits of one over the other. Voices are raised; you're passionate; YOUR idea is the best. What happens at the end? Typically, the debater with the most clout (i.e.: the client or the boss) wins. Their idea ends up getting implemented...

What's the better resolution to this dilemma (and the answer that direct marketers are typically first to suggest)? It's simple--test both concepts and let results decide the winning idea. Unfortunately, however, rigorous testing just isn't done often enough.

This article from Advertising Age, Randomized Testing Is Fast and Cheap, but Few Seem Interested made me think about testing in a new way.

It's all about real-time testing and the author cites an insightful case study of how an author let clicks from a Google ad determine the title of his soon-to-be-published book. The author was certain that he had the right title. His editor was certain that another title would be more interesting. Well, the editor was right, based on responses to the Google ad.

The article cites other testing examples and also discusses how the Internet has enabled real-time testing.

It may be time to evaluate your testing strategy. RRW has written a white paper on this subject. If you'd like a copy of that white paper, just comment on this blog and we'll send it your way. We've also discussed testing on this blog a couple of months ago. Testing is always a good idea.

Happy Monday!

Back to Basics - Direct Marketing Testing

It's been a long week (TGIF, by the way). My tiny, tired brain simply couldn't handle a slew of new ideas today. Instead, I wanted to focus on a standard principle in the world of Direct Marketing, and that is (drum roll, please)--testing. We all know we need to continuously test within our marketing campaigns. Test a new offer, a new list, a new channel, whatever--just make sure that your marketing is not getting in a rut, that you're constantly seeking new ideas and developing new strategies.

Of course, it IS easy to get in a rut--to go about business as usual. I think we've all probably felt that can predict the outcome of a test prior to implementation. And, usually we're right. If you've been doing this long enough, your gut instinct as to whether a campaign will be a success (or not) is usually on target. But, sometimes those pesky customers can surprise you. The time that the offer you thought couldn't be beat IS beat, and by an offer that doesn't seem all that attractive.

Or, when you think that the crazy new channel (like those streaming ads found in big-city elevators) won't be effective, and, lo and behold, you're getting new customers as a result of that media investment.

Just some examples of the real need to continuously test. Sometimes, however, implementing a test seems like just too much work. To spark some new testing ideas you may be interested in this article from today's "DM News": How to do direct marketing testing

It provides some basic (but valid) information on ways to implement dm tests. For example, the author (Arthur Middleton Hughes, long-time database marketing expert) gives a good definition of a control group. He also speaks to some of the measurement challenges and finishes the article with a testing checklist. All good info.

RRW also has spent some time thinking and writing on the subject of testing. Let us know if you'd like our Idea Paper, called RRW's Test and Learn Approach. It may spark some ideas of some new tests that you should implement. Just leave us a comment and we'll be happy to send you the paper.

Enjoy your weekend!