65.6% Increase In Sales So Far …
If you’ve visited the split testing section of this blog, you’ll know that I’m a great believer in split testing web pages and that I believe it’s the easiest, cheapest and safest way to increase sales from an online shop.
I’m currently running a split-test on the FAQ page of a client’s site.
I’d long believed that it was putting of prospects by focusing on questions along the lines of ”if the product doesn’t work, what should I do?”
So, I knocked up 3 “positive” questions and put them ahead of the existing questions.
Here’s a screenshot of the results so far:
(more…)
I’ve written about website split-testing before, but I want to show you exactly how powerful it can be.
Below is a screen-shot from a split-test I’m doing for one of my clients: (more…)
Q: Let me stop you and ask about split testing. What is it?
Steve: Split testing is when you’ve got an ad, then you create a different version … maybe you change the headline, but everything else is the same … so you’ve get two versions of the ad and you run them against each other to see which one pulls the bigger response.
Q: What do you mean by “run them against each other”?
Steve: Take pay per click, for example. Split testing two ads in PPC means that half the people will see ad A and half will see ad B.
After a while, you look at the results and Google tells you, “Ad A got a 2.1% click rate and ad B got a 3.2% click rate” or whatever … so you say, “Ok, ad B is the winner”.
You then create ad C, which is a variation on ad B and test those two against each other. And so on.
Q: So, you’re constantly refining the ad based on the response it gets?
Steve: That’s right, and it’s not just for pay per click, it can be done with your web pages, with your email marketing and it can be done offline.
(more…)
Split testing is one of the most powerful tools available to marketers - but one that’s overlooked by most businesses.
With the release of Google’s Website Optimizer, website split-testing is now free for anyone.
And, one of the benefits that Website Optimizer offers is that it allows you to test more than one element at a time.
For example, I have a client who has a lead generation website. We recently tested two variables on his sign-up page: headline and deck copy.
We had two different headlines and two versions of the deck copy, giving us four combinations.
Website Optimizer then rotated these four combinations until we had enough visitors (and sign-ups) to produce meaningful results.
Once the testing was over, Google gave us estimates of the difference between the headlines and copy.
Because our goal was to find the winner, rather than to know the exact difference in performance, we stopped the tests once there was more than 95% confidence in the results.
So, Google’s estimates have a certain degree of uncertainty.
Here are the results:
Headlines: headline 1 (the original headline) was 21.7% better (with +/- 11% uncertainty) than headline 2 (the test headline).
Deck Copy: copy 2 (the new version of the copy) was 37.5% better (+/- 16%) than the original.
(more…)