This Daily Golden Nugget is probably a little more advanced than most others. Thankfully, though, it's not too technical because the Google tool we're going to explain takes care of all the confusing tech stuff.
We want to introduce you to the world of website split testing, also known as A/B testing.
As we've said in the past, your website is never really complete. You need to keep working on it to improve your conversion rates, lower your bounce rates, and overall increase your sales. But other than simply watching the daily sales reports, how do you know if your website efforts are getting better?
The answer is A/B testing, and with it you can create 2 variations of the same web page and have Google run its magic until the better of the pages appears.
The magic is built into the Google Website Optimizer tool located here:
http://www.google.com/websiteoptimizer
Before you check it out on your own, let's explain how it works. Google does provide good directions, but they really don't say in plain English what it does and how it works. So let's explain it now.
The Optimizer lets you choose 3 different pages from your website. The first 2 pages are your A/B test pages and the 3rd page is the final destination page. During the test, Google will count the number of people who start at page 1 and end at page 3 and compare that to the number of people who start at page 2 and end at page 3.
In other words, page 1 is our "A" page and page 2 is our "B" page, and that's why we have A/B testing.
Now let's guide you through the steps to set this up. (NOTE: We recommend actually reading through all these steps before actually sitting down to do it for real.)
Step 1: Select the page you want to test.
This could be your home page or a landing page used in any marketing campaign. We'll refer to this as our "Page A" from now on.
Once you decide on your page, you then need to ask yourself what 1 thing on the page you want to test. It could be a headline, an image, words on a button, and even a sentence or two or the opening paragraph.
Step 2: Create another page on your site.
Use your content management system and create a new page on your site. Call it "test.html" if you like. You need to copy all the information from Page A to this new page except the 1 thing you want to test.
We're going to call this "Page B." The only difference between the pages should be that single headline, or image, or button, or that sentence.
IMPORTANT: Make sure you can view your page in a web browser BUT don't link to it from anywhere on your site. You only want Google Website Optimizer to know about it.
Step 3: Log into Google Website Optimizer.
Once logged into optimizer, you should see your "Experiment List." Google has a few different types of webpage testing methods and all together they are referred to as "Experiments." Click the link for "Create experiment" to do so.
Step 4: Select the "A/B Experiment - The simplest way to start testing fast" link.
Step 5: Verify that you are ready to proceed.
Before they let you start the experiment, they ask you to read through their A/B Experiment Checklist. It says the same thing we've already said above, but in fewer words.
Check the box to verify you've done the steps and click the Create button.
Step 6: Name your experiment and identify pages.
In this step, you will put in a name for your experiment; you could simply call it "Home Page Experiment #1"
Then fill in your Page A web address in the "Original page URL" field and Page B web addresses in the "Page variation URL" field. It will be easiest to copy the addresses from your web browser into their form.
You will also see the "Conversion page URL" field. That's where you should copy/past in the web address for the destination page.
Click the Continue button.
Step 7: Install and Validate JavaScript Tags
Okay, so we lied. There is a little bit of technical stuff involved at this step. But you can let your webmaster handle it.
On this step, you can choose whether you will install the JavaScript tracking codes, or if you will let your webmaster do it. If you select "webmaster" Google will give you a link to email to them. If your webmaster is not available, then click the "Save progress and finish later" link that is to the right of the "Continue" button.
Click the Check Status button once your webmaster has added the 3 javascripts to your website. Then click the Continue button.
Step 8: Preview and start experiment
Google provides a warning on this last setup page before the experiment starts. They want you to double check your 3 pages before the experiment starts. You should do this just to make sure it all works.
When you're ready, just click the "Start Experiment" button.
Are you wondering what happens now?
The answer is simple: Google will take turns showing Page A and Page B on the Google SERP. Your website visitors will not notice when they are sent to Page A or B; it just happens.
With every visit, Google is collecting data and trying to determine if A or B is better. You will see the report next time you log into Website Optimizer. When Google has enough data, they will tell you which one, A or B, is the better page.
When you have the winner, you should permanently commit the changes to your site and move on to the next item to test.
This is the process that all the big players use. If you've ever felt your website is not producing, then this is what you need to do to test it and make it better.
