The Secret of Writing Good Copy

    October 30, 2007

It would be annoying for me to link to grokdotcom every day, so I don’t. But if you don’t have it in your regular rotation of RSS feeds, what are you waiting for? Here’s a great example—a post that pulls together over a dozen great articles on writing winning Web copy. It’s taking some of us a while, but we’re finally figuring out that Web marketing can really depend on your words.

Direct marketers understand this. Every time you open your mailbox and get a letter offering you yet another credit card, every word in that letter has been tested to evoke the maximum response for its target market. Direct marketers don’t write one letter and send it to a million people—they write dozens of letters and send them to a small list of people to test the response rates. They send the winner to the rest of the million on the list.

Web marketing is no different, except you can do even more testing for less money in a shorter amount of time. You may not need to subject every word on every page to testing—I know that you have a life—but most companies don’t do any testing at all.

Put the page up. We’re done.

Well, yeah, you’re "done" but in a different way than you intended. You’re cooked. Your page won’t work. Because it never does on the first try.

So, yes, use all the tips you can to do a better job of creating the right words. But remember too that the most important tip is to get feedback on your customer’s response and try something else. Over and over again. Eventually you will get something that really works. (And then you keep tweaking that, too.)

So stop the "one and done." Pick just one page on your site designed to convert (no, not the home page). Pick a page that you know the purpose of—maybe the latest special offer for your top product. Come up with several versions of that page and use multivariate testing to test them all. You’ll find that some of your new ideas are undoubtedly better than what you have on the page now. Some companies report 30% gains in conversion rate after a year of multivariate testing.

So what are you waiting for?