Actually (and I mentioned this in the other item) this is exactly what I'm doing. I get a lot of repeat visitors, and I've used the same ad blocks for three years. So I'm redesigning just so I can put different sizes and colors of ads in - not messing with the content (except using a bigger font) but ads are now going to be in different places and sizes.

I wish I could come up with a way to make those big boxes work, but I just can't work it into the design.

I'm still not a fan of how the big box looks on the site but it's performing like crazy.

Sometimes money wins over style.

Funny you should mention this - I altered the layout of a site yesterday morning.

Decided a 300px box right above the content was too ugly for the earnings so removed it and added a 350px box below the content (and many of my pages are very long).

Oddly enough slightly higher earnings for yesterday than I would have expected. I guess it blends quite well now that text sizes are slightly larger, or it could be visitors are desperate to escape by the time they've finished one of my pages. Or just the effect of a change. Don't expect it to last.

I found a vertical banner to the left worked well, and when I swapped that for navigation and replaced it with a box I did notice a drop in earnings. Haven't tried a leaderboard before.

I guess if it isn't AdSense bashing or discussing poor payments then it's not worth discussing, oh well.

Which bit did I get wrong incredibill? I was trying to be constructive - I like the threads about which unit performs where as they are great for ideas. Was it the "don't expect it to last" comment? Well I don't - normally a change results in a couple of good days then things go back to where they were. I don't know why.

I believe it's worth re-evaluating different sizes and placements now that we have some control over font, and now that font sizes seem to be larger.

My big box in the footer is doing well today too - I'm hoping it will continue to outperform my slightly smaller box right at the top. But I've no idea why it should do. I've not seen eCPM like that since last year.

Anyone else experimenting? We could share ideas and become collectively less baffled.

I do like testing. I'm torn between a change I made recently. I'm still waiting it out to give it time though. I added a new block new certain style of pages, basically an index page. It has done very well, better than I ever thought. But the pages off from those index pages have lost traffic and revenue. So overall, I didn't see an improvement. I think I'm losing people to the ads on the indexes before they get to the content pages which are more targeted and probably get better ctr and ecpm and hence revenue. Still testing but getting close to removing the new blocks to see if the old patterns return and which is better.

Which bit did I get wrong incredibill? I was trying to be constructive

Sorry, that wasn't aimed at you, I was just surprised in the lack of thread participation because you would think more people would want to swap ideas that have help improve their income and learn from each other instead of just moaning about decreasing income.

I think I'm losing people to the ads on the indexes before they get to the content pages which are more targeted and probably get better ctr and ecpm and hence revenue.

Shouldn't been too hard to tell using a combo of channels and Google analytics for AdSense so you can see how many pages they view before bouncing.

Move ads close to content with 300x250 box and you will never fail. For e.g.

<h1>Page Title</h2>
300x250300x250300x250
300x250300x250300x250
300x250300x250300x250
Text Text Text Text Text Text Text Text Text Text Text Text Text Text Text Text Text Text Text Text Text Text Text Text Text Text Text Text Text Text Text Text Text Text Text Text Text Text Text Text Text Text Text Text Text Text Text Text Text Text 
300x250300x250300x250
300x250300x250300x250
300x250300x250300x250

If you have space ad text-link to left side few menu items and 160x600 block. This is what i used across all my site with good CTR.

Okay, Bill. I'm going to go test a big box, probably for a 10 day trial. I've been using a leaderboard, skyscraper combo.

See, your post is worth discussing.

Don't advertisers have the choice of which size ads they put on our sites? So if one advertiser blocks one size you use mostly, but allows another you start using, you could see your income jump.

p/g

Are you talking about image ads? A savvy advertiser (such as myself) will make sure they have ads available in ALL the offered sizes.

As far as text ads - there's usually several to an ad block; advertisers have no control over that.