Make More Money by Analyzing Your Adsense Stats
I am something of a stats head and can easily spend hours going over traffic graphs, Adsense screens and referral logs. One of the great things about the web is the shear amount of stats available, you don’t need to employ mystery shoppers or send out surveys to your customers, you can simply install a script and find out more information about your visitors than you’ll ever need.
One thing you need to be cautious of is checking stats for the sake of checking stats. Like most webmasters when I first started out with Adsense I’d be checking income every hour or so to see if there’d been any clicks, a complete waste of time but difficult to resist. As income goes up it does get easier though I still find myself checking at least once per day, especially when approaching milestones.
On individual item pages (example) I have two units, a 468×60 at top and then below that after the title of the post a large 300×250 rectangle (see diagram). What I did was check the CPC and CTR of the two units, I can’t give exact details as that would be in breech of the Google TOS that I agreed to however I’ll share the comparison between the two.
The CTR of unit 2 (300×250) is over 10x that of unit 1 (468×60), whilst the average CPC of unit 1 is ~50% higher than that of unit 2. Which means the unit that gets most clicks (by a huge margin) earns less per click, so if I could reverse that trend, overall income would go up without needing more clicks. So time to do a test.
It’s a common believe that the first Adsense unit on a page (when you view source) gets the highest paying ads, which makes sense as you’d expect the most profitable advertisers to appear first. So with that in mind I used CSS to position the units in exactly the same place as before but have unit 2 appear first in the code (if you want to see the code to do that just let me know). I then needed to track the results and see if the was any difference in CTR or CPC.
For tracking changes this is were being able to have more than one channel associated with each unit is a real blessing. All you do is create 2 new channels (say Position Test 1 and Position Test 2) and then modify your sites code to use each method 50% of the time and append the appropriate channel to the Adsense code. I let it run like this for a few days and a couple of thousand clicks and then compared the results.
To compare the results the easiest way is to create a report that only shows your new channels and export it as a CSV. You can load load it in a speadsheet and accurately compare the results (adding a field in to view average CPC helps (+Earnings/Clicks)). If you prefer looking at graphs to figures you could also use a service like the DP Adsense charts.
After doing this little test my earnings are up by just over 4% which isn’t as much as I was expecting but not bad for the small amount of work involved and over the year this will hopefully be another $5000 hour.
You can also use this method to test different color schemes, unit sizes etc. which I’ll be doing over the next few weeks.
Related Posts:
I would have expected a higher increase although at your level of income the 4% is still nice. Have you tried different ad networks on the top spot to see if that helps increase the lower units?
Also, how did the eCPM compare after the switch over? did you see a decrease on the top unit and an increase on the top unit?
I’ve not tried another network in the top spot yet, even though the CTR is relatively low the Adsense income from that spot is still quite nice which I doubt any other network would match, though I will have to test it to see how it effects the other unit.
Overall the eCPM of the top spot reduced slightly whilst the eCPM of the rectangle increased about the same but with the extra clicks in that position overall revenue was up.
I have several adsense channels set up to cover all my sites I use adsense on so I can see which sites are performing and which aren’t. On my forum I use several channels for different positions so I can which ad position works best.
I only really look at overall income though and have never really done much analysis beyond that. For instance my forum gets loads (well, compared to my other sites) of impressions but I can often get the same amount of income from two clicks on another site as 20 clicks on my forum.
I’ve not really looked into ways to improve things…. maybe you could do a few more handy hints for running adsense to help us low earners improve a little.
I’ll certainly put my thinking hat on for that one, I also find forums are notoriously hard to monetise with Adsense but if you can use the forum to promote another part of the site that is easier to monetise then that can work.
Yeah, my forum has proved hard to monetise. Now the membership is reasonable, I have expanded the site so I can add articles and other pages which may be better for adsense and similar.
This is a very helpful post! Not many people actually talk in detail about these types of tests but I’ve always wondered how many people actually test these things vs. going with assumptions.
You can learn an awful lot reading forums and blogs but nothing beats doing it your self, every site and niche is different so what works for one person may not work for another.
Do you think if you replaced Ad block 2 with two rows of 468×60 and appear first in the code using CSS that it would give you the CTR and the better CPC, I have seen it done plenty of times but never tried it myself as a 300×250 seems better but maybe not if the resulting CPC is lower.
That’s an interesting idea, I must admit it does scare me playing with one of my most profitable bits site real estate but it’s probably worth trying.
“First Adsense unit on a page” has new meaning with CSS. This and the SEO trick of putting your main keywords higher in the HTML, regardless of where they are on the page, is just a bit funny. I hadn’t thought of using the trick in terms of Adsense, thanks!
I added this page to my favorites for later use. Interesting way to boost your earnings.
Have you tried to remove one of the blocks completly? Therefore serving less ads overall but hopefully of higher value. Maybe worth a test for a few hours (don’t want to lose you too much money if it goes wrong
That is certainly another thing I could try. What I’ll do is create a few test cases from the suggestions in these comments and run them concurrently for a couple of days and report my findings. The pages in question generate ~$200 a day from Adsense so the most I can lose is $600 but hopefully I’ll find a way where I can be making more.
I believe ad block #1’s cost per clicks are higher because the ad unit appears higher in your source code. Google prioritizes the ads it places on your site like that, so if you could get ad block #2 to appear higher in your source code, you could combine the higher CTR and CPC. A winning combination.
That’s what this post was all about Ed and I did move Ad Block #2 to be first in the source code, it did improve the CPC but no where near to the same level as Ad Block 1
SMM, sorry I missed that. Interesting test.
Could you post this?
I’m not sure how the code will appear in these comments but I’ll have a go:
In the main html/php file I have:
<div class="SingleAdTop"></div>
This is just a spacer where the image will appear.
Then at later in the code I have the Adsense script and use CSS to put it in the gap I previously left:
<div class="AdPosition">
<!-- google_ad_code -->
</div>
Finally in the CSS I have two new classes:
.SingleAdTop
{
height:60px;
width:468px;
}
.AdPosition
{
position:absolute;
top:235px;
}
Make sense?
Yeah, thanks alot!
Well I hope I was still on adsense to do all this
Yet another excellent article
I think you may be going about this the wrong way. The best way to optimise your earnings will probably be to remove the 468 altogether…..because the high number of impressions with a low CTR is bringing down the overall value of ads that google is sending to your site.
If you remove the 468 which will improve your site’s overall CTR, will mean that google will give you the higher value ads because of smart pricing, which will also be displaying in your unit with the best CTR.
May seem weird that removing a unit will improve overall earnings, but it has worked for me and many others.
Thanks Everton, I have experimented with different number of units before with various levels of success. How long do you normally need to leave it to see the full effect?
took a few weeks for my cpms to adjust
Thanks Al, an interesting suggestion and proposition, and thanks for sharing the code.
Thanks Al thats a great tip
Everton Thanks that could explain some things i have going on.