Using Split Testing, A/B Testing, to Optimize Adsense Earnings

If you don't you're leaving money on the table.

[This is part of the Blogger’s Guide to Earning More with Google Adsense.]

Well, we’re finally here – one of the “big” posts about Google Adsense testing that I’ve been holding off so as to give more context for what we’d be doing and prepping you for the work that would be involved.

The fact is this: There are tons of optimization techniques that exist (and I’m not going to share all of them in this series, be certain of that), some of which I’ve already shared but there are few better than “Split Testing” or what is also understood commonly as “A/B Testing.”

What is that? Most simply it’s this:

A/B testing, split testing or bucket testing is a method of marketing testing by which a baseline control sample is compared to a variety of single-variable test samples in order to improve response rates.

What you are simply doing in the context of Google Adsense is seeing how your advertisements perform in different situations, locations, size, etc. In fact, all of the optimization techniques that I’ve shared so far can be accurately measured by the A/B Testing technique!

Here’s a very practical example:

  • You begin by adding a 468 x 60 banner advertisement in the header for two months. It brings an average of 10 clicks per month for you earning you about $20.
  • You move the same sized banner into the content layer right before the title of the post. You test for the same amount of time, two months. It brings an average of 15 clicks per month earning you $30 per month, an increase of $10!
  • You opt for sticking with the 468 x 60 right before the title from now on (or you could test another location, placement, and size).
  • Rinse and repeat.

This is a very simple (yet realistic) example of a basic split test or A/B test. Here’s what you will need to be successful at this:

The 3 Requirements for Good A/B Testing:

Generally you’ll need the following so that you can be successful at split testing.

These are not necessarily hard and fast rules (or requirements really, although I used that word…) but will increase the likelihood of you getting accurate and dependable results:

  1. Traffic – I’ve already shared some thoughts on what it might take traffic-wise to get Adsense up and running but you probably need a bit more to start split testing effectively. In fact, I’d recommend that you spend your time elsewhere until you have at least 50,000 pageviews per month, consistently. 100k is probably better and which is why I have waited nearly a year to start this type of process myself (and it’s only been 2 months since I broke 100,000 pageviews). There just wasn’t enough reason to even begin until I had enough traffic to play with.
  2. Time – One of the biggest challenges that most people have with split testing is their impatience with the process. You have to give the tests ample amount of time to prove the differences you are attempting to analyze and most bloggers (and many people in general) aren’t willing to wait to see the results and will test for a week (or two) before making a decision. You have to overcome the temptation to jeopardize the completeness of your test for getting faster results. In addition, you need time to create the tests, administer them, and poll the results.
  3. Test Cases (Simple) – What I mean by this is that you simply need simple and straight-forward test cases that can provide a reasonable chance of creating a meaningful set of data to help you optimize your efforts. The more simple the better, especially in the beginning. You can create more elaborate tests down the line but you may never end up having to do that since I believe the most simple tests can prove to be the most profitable, and your best ROI.

I could probably add a few more things here but that should put a lot of things into perspective. In fact, I’ll give you another practical example of why you should really opt to wait and spend more time on creating amazing blog content until you have the right traffic:

10k Blog:

Consider the following example:

  • Let’s say you use the “Blogger’s Adsense Combo” on your blog.
  • Your blog has about 10,000 pageviews per month.
  • Your total CTR is 2%.
  • You CPC is $.20.
  • Your total monthly earnings will (have been) $40.00.
  • I get this by the following equation: PVs x CTR x CPC. So that means: 10,000 x .02 x $.20 = $40.00
  • You spend time creating tests, recording data, changing the flow of your blog, even the design, optimizing it for traffic, moving ad space around and you move from a CTR of 2% to 2.5%.
  • You now earn $50.00, an increase of only $10.
  • You think of all that work that was put in that could have been spent on creating amazing blog content, curating (and expanding the community) and sharing content around the web.
  • Bummer.

Now, let’s consider another scenario…

100k Blog:

Now, you do the same things that were done on the 10k blog and the results are dramatically different, especially in terms of the amount of time spent:

  • You have 100,000 pageviews per month.
  • CTR is 2%.
  • CPC is $.20.
  • You make $400.00 per month.
  • You optimize, move to CTR of 2.5%. You now earn $500 per month, an increase of $100.00!
  • Time well spent, win.

This is a far better scenario in my opinion since you want to maximize your time always with your blog.

Google’s Website Optimizer Tool:

The next step is actually creating the tests and the first option that many people suggest is to simply use Google’s Website Optimizer tool:

Here you can create A/B Tests to your hearts content by creating new “experiments”:

It's fairly straight-forward.

Simply choose the easier option to begin with:

Don't go crazy on it yet...!

Then all you have to do is walk through their wizard:

Click, click, click...

I’ll admit that although it’s fairly straight-forward it may take you some time to get adjusted and things in place to make it happen.

It’ll eventually spit out a report for you that you can use to optimize your earnings!

DIY Split Testing (My Preferred Method):

The other alternative is to do it yourself, which is what I typically suggest the new optimizer to try first before using an automated tool, simply because I think it’s more quick and easier to do.

All you have to do is use your current Adsense Control Panel and track the numbers that you need in an excel spreadsheet over a period of time that you control.

Then, make the changes for your test and wait for the same amount of time. Take the results of the second (“B”) test and compare them with the first (“A”) test. See how you do and then make a decision or test some more.

Done and done!

Again, I think this option is a far-simpler option to move ahead with first before using any fancy tool. Get used to the dynamics of testing first before automating it.

Finally, there are a ton of other tools out there that can help and you can do your independent research to find them but I’ll admit that I still do manual testing myself and I’ve done quite well with it. To each his own!

Good luck and let me know how it goes!

[This is part of the Blogger’s Guide to Earning More with Google Adsense.]