I thoroughly enjoyed my first trip to Winnipeg, speaking to a couple of hundred marketers at the Canadian Marketing Associate Digital Days conference. (You can download the slides for my talk, “How Web Marketing Changes the Old Marketing Rules.”) I spent lots of time hammering away at how to do it wrong quickly, how to experiment in marketing, and how to measure the results. And someone from the audience asked a great question, “Can experimental marketing alienate your customers?”

It’s a great question, because there are things you can do that would be horribly alienating. If your insurance agency is considering a titillating ad to run on YouTube, I might say, go slow. There are definitely risky behaviors in Internet marketing that might not be your best first experiment. But most things you do are not so risky. You can try them and take them down if they don’t work.

When I was at IBM, I remember testing something in Russia to see if it worked, because testing it in the US would have gotten too much attention to reverse if it was bad. You can usually find a way to test things under the radar.

But what if you really do alienate people? What if you screw up and do annoy someone? Or even a lot of someones? The way this plays out depends a lot on your response. If you just take the offensive material down and pretend it never happened, don’t expect people to be impressed. You might find the conversation to be decidedly negative and perhaps even grow worse, all because you are not responding, so it looks like you don’t care.

On the other hand, you might apologize. You might answer as many of the negative comments as you can, no matter where they are found on the Web. You might continue the conversation so that you really understand what ticked people off, so that it truly will never happen again. If care about your customers’ feelings more than your own, it will show. And you’ll end up alienating very few people.

Next time you screw up, pay attention to what you do next to mend the relationship rather than running away and hiding. Then try your next experiment now that you are a little smarter.

Comments