The idea of A/B testing has long been used with designers but definitely has applications with online marketing as well. It was my friend Lynn Cyr, a user interface designer, who introduced me to the concept via a blog.

A/B testing is simply testing two versions of the same email newsletter, landing page, contact form, etc. and see which one performs better.

Here is a simple example from CampaignMonitor.com (click here to get the full post from the site):



There's not much difference in this email campaign except the words used in the link. Guess which one did better?

There’s not much difference in this email campaign except the words used in the link. Guess which one did better?

So I am going to kill the suspense here: Campaign A had a higher response rate. Here’s the graph for those of you who like a visual:

You can lie to yourself about how effective something is but your numbers won't lie for you.

You can lie to yourself about how effective something is but your numbers won’t lie for you.

And you may think, great, nerds the world over can figure out how more people can fill out a contact form. What does this have to do with me?

What if I told you you could test a new website configuration to sell more products?

ABTests.com has visual case studies you can learn from and be inspired to try yourself.

ABTests.com has visual case studies you can learn from and be inspired to try yourself.

In the example above, adding a ‘Popular Products’ feature increased sales with those products while people were looking at related products (AKA cross-selling). This example makes sense but not all of the case studies seem logical. Here’s one with interesting results from ABTests.com:

You'd think the cell phone would speak better to active technology saavy folks on the go (at least I would!) but it didn't. That's why we test, right?

You’d think the cell phone would speak better to active technology saavy folks on the go (at least I would!) but it didn’t. That’s why we test, right?

Here’s the thing: As business owners, we can not assume we know what our customer thinks. Scientists run experiments and psychologists conduct studies precisely because, while we can come up with a prediction, we always have to test it. But besides the quest for knowledge, why A/B test your Facebook page, website, email newsletter, blog, or anything?

1) More sales with less work.
Let’s say you spend an hour setting up Version A and Version B of something. Wait a month while data collects. Look at the numbers, do what works. Even if you only increased sales by 5%, that’s more money for the same amount of effort, that means more money for the same amount of website traffic you are already getting.

2) It’s not expensive.
If you want to DIY, companies like Visual Site Optimizer offer a 30 day free trial (and less than $30/month after for small businesses) which allow you to set up your own tests and track the results. Or you can have your favorite nerd set up something for you. Even if you pay a nerd like me $75/hour for 1-2 hours work, you’ll more than make that money back with your improved results.

3) It means better results over time.
So you find out your customers like one website layout over another. Guess what you’ll consider when you redesign your email newsletter? Exactly. A/B testing means you get better at reaching your target customer over time, further improving what you are trying to do both on and off your website.

Your website visitors can tell you so much without saying a word. And while we’d all prefer to think our websites are perfect, we can always do it a bit better. A/B Testing just helps you get to that better level quicker.

For more information about A/B testing including setup, check out this article from Smashing Magazine.

Or contact me if you’d like me to set up an A/B test for your contact page, newsletter, Facebook ad, or whatever!

Need marketing help?

X