The Conversion Rate Problem


June 16, 2007

Every day, or maybe every other day, we get a note in our email that usually take this format:

Hi LunaMetrics. Our conversion rate sucks is just terrible. What can you do to help us?

I don’t mind getting emails like this; after all, the potential customer is just starting a dialog that might lead to a big project. But I feel that it’s important for the potential customer to understand how much effort might have to go in to increasing a conversion rate.

Any more, I’ve starting to answer those emails with one of my own, and it goes something like this:

Let me not steal your money. Sure, we’d be happy to set up your analytics, and then collect enough data to read them and make recommendations, if you need that. We can do some user testing to find out what customers think (always incredibly valuable) and we can even do a conversion best practices analysis to tell you what we think. And we’ll have a great idea of why you aren’t converting well. But I don’t want you to think that we’ll do that analysis, you’ll make a few changes, and your conversion rate will magically increase. It requires testing, testing and more testing.

At the first Emetrics Summit I attended, in Santa Barbara, I heard Avinash Kaushik say, “80% of the time, we are wrong about what customers want.” Only by asking them – user testing, multivariate testing – can we see what they really want.

And then, let’s look at this problem a different way: when the problem is not the site, but rather, the position. I’ve seen too many sites that sell commodity items without a unique selling proposition. So after we do that user testing, we’re going to find out, for example, that users don’t know who you are, and they can buy your product for the same rate elsewhere, so why should they give you their credit card numbers?

This doesn’t mean that all commodity businesses have lousy conversion rates. On the contrary, one of our customers has a website to sell their commodity; their conversion rate would make your skin turn so green that you could get the part of Elphaba on Broadway. But instead of starting with, “What color should our buttons be?” or “What should the headline say?” they have consciously thought through the offer and the barriers to purchase, and have done everything they can to make this the best offer and easiest purchase you could ever want to make.

So before you write that email asking to increase your conversion rate, think about whether you have a unique selling proposition. And then, before you waste a dime, decide whether you are ready to put in the work to test and test and test in an effort to increase those rates. Because otherwise, I really do feel like I am stealing your money.