Welcome to WebmasterWorld Guest from 18.204.227.250

Forum Moderators: martinibuster

Message Too Old, No Replies

Experimentation Proves I Should Stop Trying So Hard

     
3:05 pm on Aug 24, 2007 (gmt 0)

Junior Member

10+ Year Member

joined:Apr 10, 2007
posts:41
votes: 0


I've only had Adsense on my site since Sept. 2006. My earnings dropped through the floor this summer and I suspect that's going to be a regular seasonal fluctuation.

Since I didn't have much to lose anyway, I experimented all summer. I've tried many of the tactics talked about here, some of my own, and none of them have worked. As a matter of fact, they made things worse.

Here's what I tired:

Blocked lots of sites that seemed spammy and mfa - my price per click plummeted and my ctr plummeted, too.

Removed an ad unit from each page - this didn't increase the value of my clicks, it just reduced the number of clicks.

Replaced removed ad unit space with a Referal Ad Unit - Zero clicks out of several thousand impressions.

Replaced removed ad unit space with an affiliate link - Maybe one click out of several thousand impressions. No sales.

My 120 x 90 link unit was getting a lot of clicks so I replaced my four link unit with a five link unit - ctr for that unit fell dramatically.

The last thing I experimented with was taking Adsense off the pages my own products are sold from. I don't buy things from people who have Adsense on the same pages as their products, I've heard many of you say you think that looks cheesy, so it made sense to remove those ads. Well, I sold fewer of those products than ever before and I missed out on the extra Adsense revenue, too.

I was very scientific about these experiments. I tried one thing at a time and collected statistics for several weeks after I changed something.

I've put the link and ad units back to the way they were in the Spring. Things are picking up. Life is good. I've learned a lot.

My visitors, bless them, like big fonts, lots of Ads, and especially can't resist those ads for sites that offer a free product if they follow a "reward path". They are perfectly willing to buy my products if there are Adsense ads on the pages that sell them, and will probably click an ad if they don't.

I love my visitors. I can kick back and be very lazy about all this. No more tweaking. I'm just gonna write lots and lots of good content and plug it into my template, which is what I really love to do anyway. :)

3:19 pm on Aug 24, 2007 (gmt 0)

Full Member

10+ Year Member

joined:May 2, 2006
posts:243
votes: 0


IMO "several thousand impressions" is not enough data.

For example, when you block ads (for whatever reason), it takes maybe even millions of impressions for adsense to really learn what kind of ads from thousands ads available perform well on your pages.

Let's say you have 100 pages with adsense. And you test something new with 100,000 impressions. That's only (100,000/100=) 1000 impressions per page. Let's say there are 100 possible ads for each page, it means (1000/100=) 10 views per each possible ad. That's clearly not enough views for adsense to really know if that ad is good for your site.

_

3:36 pm on Aug 24, 2007 (gmt 0)

Junior Member

10+ Year Member

joined:Apr 10, 2007
posts:41
votes: 0


Hmmmm. Well, I should have said several tens of thousands of impressions. And statistics being what they are, and Google being who they are, I think that would be enough data to be pretty darned conclusive.

This is what works for my site, my niche, and my visitors. I'm perfectly willing to be an anomaly.

4:35 pm on Aug 24, 2007 (gmt 0)

Preferred Member

10+ Year Member

joined:Aug 15, 2003
posts:421
votes: 0


You are right on target, imho. Good content -- and lots of it -- is what brings visitors to your site, keeps them on the page and maybe even brings them back sometime.

We have had AdSense on our site since Day One and have achieved consistent results, except for a few weird periods, over the years. We add content constantly -- reader comments, our own writers' efforts and a little bit of syndicated copy -- and fiddle very little with the AdSense ads. We are always trying other ad networks; some are OK but nothing even comes close to the Big G.

5:53 pm on Aug 24, 2007 (gmt 0)

Preferred Member

10+ Year Member Top Contributors Of The Month

joined:Mar 12, 2004
posts:515
votes: 25


I've been doing the same over the summer despite it having been proven in the past by Mr 21blue (where did that dude go?) that my efforts are statistically insignificant.

Same result as yourself. It's almost as if the big G has decided how much they'll give me for each page. I ended up just altering things to make the site prettier and for me that didn't reduce the G-money.

Though my niche is small and I can only get what the advertisers are paying (less the stuff MFA are getting and the G cut).

7:12 pm on Aug 24, 2007 (gmt 0)

Junior Member

10+ Year Member

joined:Apr 10, 2007
posts:41
votes: 0


It's nice to hear confirmation of my conclusions from people who have a lot more experience than me. Thanks for that.

The main reason I posted about this is that it surprised me and I'm still trying to digest it! After several months of working on tweaking my ads, I'm right back where I started. My visitors come to my site to read my work. That's what brings them in and it's the reason they stay. They don't mind ads; they even seem to like them.

And that's fine. It's great, really. Now I can go back to writing content without any nagging suspicions that I could be earning more if only I could get the ads right.

I've learned tons and tons in the process, and I thank you all for sharing your knowledge. Most of all, though, I've learned where and how to focus on my site to make it even more profitable and that's right where I started in the first place. I guess all good journeys begin and end at home.

Click your heels together three times, Dorothy. :)

7:55 pm on Aug 24, 2007 (gmt 0)

Junior Member

10+ Year Member

joined:July 27, 2007
posts:76
votes: 0


I think most of your visitors are loyal so they got familiar with the ads.

I think also that adsense works nice with a site brings most of the visitors from search engines.

8:38 pm on Aug 24, 2007 (gmt 0)

Senior Member from GB 

WebmasterWorld Senior Member 10+ Year Member Top Contributors Of The Month

joined:Apr 29, 2005
posts:2112
votes: 122


""I've only had Adsense on my site since Sept. 2006""

In that time you have conducted many experiments. Looking back at the experiments I have done on my sites, I've noticed that real results or failures are only apparent after 3 to 6 months. In less than one year of Adsense it may be that you have not allowed your changes to realise their true results before another change has obscured if they have failed or succeded.

Of this I am sure (many times over), if you make a sitewide change, over the next couple of weeks you will see a significant improvement or decline in earnings. I have no idea why. But in the months following the change you will then see the real result of those changes. Some are good, some are bad. Significant changes to a site take several months before you know if they are good or bad. And this advice comes from someone who is earning from my own sites and not from an SEO optimiser. It does take a few months before you can judge if a change is good or bad.

9:02 pm on Aug 24, 2007 (gmt 0)

Junior Member

10+ Year Member

joined:Apr 10, 2007
posts:41
votes: 0


I thought about ad blindness, too, and it was something I was worried about for a while. I tried blocking a few advertisers who park permanently on my site just to get different ads for a while, and I've come to the conclusion that I'm a lot more tired of seeing those ads than my visitors are.

I will mention one more thing I did that was informative and interesting. I put a little pixel on each page from a company who looks at visitor's demographics. That was really useful to me. Learned a lot about my visitors that I never would have guessed and now I'm using that info to help guide the development of some areas of content.

9:20 pm on Aug 24, 2007 (gmt 0)

Junior Member

10+ Year Member

joined:Apr 10, 2007
posts:41
votes: 0


if you make a sitewide change, over the next couple of weeks you will see a significant improvement or decline in earnings. I have no idea why. But in the months following the change you will then see the real result of those changes.

I believe you. I don't have a lot of experience at this. My conclusions are based on my limited knowledge of how statistics work, and I'm assuming that a "small" sample can represent a much larger trend. It's not that simple, though.

Based on what you say, I'm never going to be able to test a change completely because I don't have that kind of patience! So, I'm gonna go with what I've got. And I'm a lot wiser from the effort.

10:13 pm on Aug 24, 2007 (gmt 0)

Senior Member from GB 

WebmasterWorld Senior Member 10+ Year Member Top Contributors Of The Month

joined:Apr 29, 2005
posts:2112
votes: 122


Forest

You've got it right. Go with what you know, be aware there is a big picture and work into that big picture as time goes by. Good luck.

1:50 am on Aug 25, 2007 (gmt 0)

Junior Member

10+ Year Member

joined:Apr 5, 2003
posts:96
votes: 0


Forest, seasonal traffic variation in my case is about 50% down in summer. A lower CPM amplifies the reduction in earnings.

About content I think you are correct. Thanks for remind it to me.

5:21 am on Aug 25, 2007 (gmt 0)

Senior Member

WebmasterWorld Senior Member 10+ Year Member

joined:May 9, 2007
posts:1362
votes: 0


Forest:

I think you are quite right about your conclusions. Also the size of the data set seems appropriate to me. Only recently I came to the conclusion to let Google do its work and serve ads without my interference. Though I really hate to see all those parked pages advertised through my sites (well knowing that someone out there makes a living from doing virtually nothing while we have to pay for content development) I stopped blocking this crap because it's too much effort to find the new parked sites. Also, it did not increase the payout or total revenue. So my visitors have to live with the fact that Google likes to sell ads to crap companies. Well, that's the way it is. So I can lean back and focus on content development.

Anyway, excellent post.

7:48 am on Aug 25, 2007 (gmt 0)

Senior Member

WebmasterWorld Senior Member 10+ Year Member

joined:Apr 1, 2006
posts:1536
votes: 0


Forest

IMO You got most of it right.

It makes no sense to keep on tweaking. Experiment a bit and then stabilize things. Adsense works best in a stable environment. As Zett says it takes a while for G to find the correct ads and they do get it right most times. Some MFA's can't be avoided.

Adsense on e commerce site is hotly debated on this forum. What works with some does not work with others. For my e commerce site, I feel adsense adds value by providing other avenues for surfers as I work on a very specific niche. Some surfers see adsense as a measure of ' professionalism' and 'trust'. They have told me so.

Adding more good content ensures repeat vistors, so that is the right way to go.

Best of luck..