eHarmony is launching a 10th anniversary campaign. They are not doing the testimonial thing, nor are they hawking their profiling feature. Good. There is no evidence that it produces better results than any other method, including letting people do their own screening. Rather, they are running warm fuzzy spots with the likes of couples swinging on a hammock and kissing underwater. Given the strong association between a warm fuzzy and romance, which is ultimately what eHarmony purports to sell, the campaign might just work.

Still, I had to laugh at this gem from CEO Greg Waldorf: “People don’t want to be sold to anymore.” If he were my marketing director, I would fire him. The statement is provably untrue, and is used by marketing people to justify wimpy, indirect, ineffective work. Unfortunately, this myth flourishes among marketing people—most of whom, ironically, have never sold anything for a living—which is why so much marketing fails miserably.

Plus, his use of “anymore” begs the question as to whether there was indeed ever a time in which people would have expressed a heartfelt desire to be sold to.

The question is not whether people like to be sold to, but how they like to be sold to. A look at what most often leads the most people to buy shows that, more than you might expect, a hard-sell approach is exactly how people want to be sold to. They will tell you the opposite in a survey, but their actions show otherwise. Not that hard-sell is the only viable option. Sometimes people want a calmer, reasoned approach. Sometimes they want a straight, factual one. And, yes, sometimes they want an emotional one, of which warm and fuzzy is a variety. How do you know which to use when? Set up an empirical test and then count the sales.

More often than not—and this is not easy to get across to clients—the kind of marketing for which people express a preference, and that clients are most proud to run, is the kind that doesn’t make an iota of difference in sales.

Do not mistake what you personally like and dislike, or do and don’t respond to, for the norm. For one thing, you alone are not a statistically valid sample. For another, you haven’t a clue as to what you respond to. Rather, you know what you’d like to think you would do, based on your self-concept. I’ll never forget the corporate decision maker who heaved me out his door because he never, ever responded to direct mail, much less my style of it, and that his customers were just like him. Shall I tell you what he didn’t know? That I mailed to him regularly on behalf of one of my clients, and that he was their most frequent respondent. This is why smart marketers test rather than assume.

—Steve Cuno