• Home
• Press Room
• Testimonials
• Code of Conduct
• Tips & Tools

> Advertising > Ad Update

Using A/B Split Testing to Maximize Your ROI

Why Should You Test?
Imagine. Your business gets 600 visitors daily, your conversion rate is 1%, and your average sale is $40. That would put your annual gross sales at about $87,600. Now, say you bumped your conversion to 2%. Just a slight 1% increase in your conversion rate would DOUBLE your sales volume to over $175,200!

Why test? The obvious answer is that customers vote with their wallet. And frankly most entrepreneurs are so wrapped up in their product, service or offer that they think they just “know” what will be the best performing offer, piece of creative, or headline or landing page.

What we really need to do is to put our assumptions aside and let the numbers tell us what reality looks like. If you’ll take that one idea and put it into practice you just may find a sizable increase in results… whether it’s more sales, more customers or more subscribers to your house list. Any and all of which would be a good thing!

A second reason to test is that now it’s easier than ever. Even smaller advertisers, mom and pop shops, members of affiliate programs, etc., can use simple A/B testing strategies to great effect and real success. Your benefits could be higher click through rates and on the back-end increased conversions. Sure, the complete process will take some time and willingness to stay the course, but once you have the testing plan clearly defined and laid out, it’s well worth it.

What To Test?
For our purposes in Newsletter Advertising, using banners that ride along with the content, we recommend testing the following four areas: offer, ad creative, headlines and landing pages. You may start with just one area, then move to additional tests as you progress.

Keep in mind that combining “winning” four areas together may not produce the best result. For example, after completing isolated tests of each campaign element, you pair the “winning” ad creative and landing page. Although the clicks to the landing page remain the same, the conversion ratio of clicks to sales dramatically decreases. Why? You made the mistake of assuming that the best banner ad and landing page work together. In this case, the banner ad and landing page must work together to create a cohesive user experience.

Offer -- When testing your offer, vary several offer elements one at a time. Offer elements could include: free trial, free shipping, discounted price, free bonus with purchase, or soft /hard sell approach. Test and find out which offer works best!

Ad Creative -- When testing your ad creative, consider varying the graphic elements, color choices, backgrounds, sizing and types of font, and image placement in relation to the copy. Should you use a button or a simple link as the call to action? All of these elements can significantly impact the click through rate of your creative.

Headline -– Marketing veteran, Ted Nicholas, has said for years that, “The headline is the ad for the ad.” If your headline is weak, fails to grab the reader’s attention, engage them emotionally, and pique their curiosity in some way , your audience will probably gloss over the rest of the banner. Remember that only your headline should change during the headline A/B split test. Everything else stays the same.

Landing Pages –- Testing Landing Pages garnered a 73% response in the MarketingSherpa study for those noting a “significant difference.” Your conversion ratio of clicks to sales (or other desired action) is what you are measuring here. You might test long copy verses short copy, multiple call to actions verses just one, or the impact of testimonials. Your publisher should be able to use two different links for the same ad creative for the A/B split test. Or, use a rotator link to automatically rotate visitors to two different landing pages for your test.

A word of caution… Test only one variable at a time. We’ve made the mistake of trying to pack too many variables into a single campaign. This practice muddies the water and your results will not be definitive. You’ll end up asking yourself, “What worked? Moving the before and after photos above the headline, or taking out the back ground color?” And you’ll have to repeat the test.

Getting your tests set up and running should be fairly straight forward. Submit two offers to your publisher and ask them to split your distribution. Many publishers will offer the service at a nominal cost or no cost if your buy is large enough.

Measure Results and Keep The Winners If you’ve been marketing or advertising your offer for a while, then you should already have some very valuable data that can serve as a control. Measure your tests against that data. Be clear in what you plan to test and keep detailed records on your results. Define the variables for each test and chart the variables and results to analyze your results. Determine in advance what will be considered significant difference.

And, don’t forget to be realistic in your goals. While a 75% increase in the click through rate would be a colossal achievement, it’s not likely. A 5% increase in clicks or sales conversion over time can be extremely valuable. Just imagine the increase to your bottom line over time. You'll keep the winners and build future tests from there.

Keep Testing Be committed to the process and you’ll find the optimized combination of offer, creative, headline and landing page. Should you quit? Never. There’s always room for improvement. Keep your “control” or best producing combination in place, but stay at it! Run a new test. Analyze the results. You may find a new “control” and yet another increase in conversion and revenue. Then test again!

Do you have questions about testing? Email ArcaMax at advertising_advice@arcamax.com and we'll do our best to provide meaningful advice and insight.



This news arrived on: 02/26/2008

---

Printer Friendly Version | Send this page to a friend | Post Comment

Rate This Story:

Great - 5 - 4 - 3 - 2 - 1 - Bad



Thank you for your input.