Sorry, I don't have a "yes, and the details are Here: ..." answer, but maybe some policy exists - loosely related to their AdSense "optimization" suggestions to load your own page's js and css after the page loads. I do not see how that could please anyone viewing the page, just so their ads could load first. IF there is such a policy you can find it in AdSense help links.

Well the last time I checked the answer was you weren't allowed to make modifications to the ad code. However, now I'm seeing something different:

Ad behavior

Publishers are permitted to make modifications to the AdSense ad code so long as those modifications do not artificially inflate ad performance or harm advertisers.

[support.google.com...]

So, I have no idea but if it were me I'd go with what the AdSense rep said.

How are you going to code the ad? If you're only going to wrap it with div class, then you are fine. Just make sure you don't modify the original code.

I was not planning on modifying the code itself. Just wanted to make it so it did not load until someone would actually see it.

Maybe this part is where it might be the issue:

so long as those modifications do not artificially inflate ad performance

In a way, this would artificially inflate the ad performance. I did say, I was looking to improve my CTR by not displaying ads that normally would loaded but not be seen. But, the advertiser would also benefit because I am sure they don't want to pay for ads to be loaded that don't get seen. But, by the strict interpretation of the rules, that line does apply regardless of who benefits.

I'd err on the side of caution. Maybe someday they'll relax that, but for the moment, it sounds risky.

I'd err on the side of caution.

I always do. I am just frustrated by this. We were also talking about adding infinite scroll, but now I am freaked that it may be a policy violation as well.

oops, wrong thread. sorry :D

I don't think infinite scroll is a policy violation, but I've heard that Google may not index everything in that situation, so you might want to keep that in mind.

I wonder if waiting until a reader scrolled down to an ad before it loaded wouldn't impact earnings. Doesn't Google count a page view when the ad is loaded (not necessarily read)? If that's the case then you could lose revenue.

WDR

If that's the case then you could lose revenue.

Yes, this is the case, but I am a big believer in testing. My idea would be that I would hopefully be cutting off the bottom of the barrel CPM ads for more lucrative CPC ads. But I would have to test to see if it would actually work.

Then there is another issue. Google would also lose revenue (they do charge advertisers for page views) and might not be pleased about that. (Although G is supposed to bend over backwards to make sure advertisers get value for their money and the advertisers would probably love it.)

In fact, if I were an advertiser I'd be overjoyed if every single publisher made the best effort to ensure that they only shelled out for a "view" when it was pretty certain that a visitor had actually seen the ads.

WDR

So, the policy team did get back to me and it is indeed a policy violation to lazy load ads. So I guess I won't be doing that.

But, I was curious to see how many people actually did get to see the ads loaded at the bottom of the page. So I asked my dev to put in a simple script that would count how many people made it to the bottom of the page.

The result was just shy of 7%.

The reason that I am sharing this is because it is not a metric I had never thought of before. We all look at things like time on page or bounce rate to judge how well we are interacting with our audience but seeing how many people actually make it to the bottom of the page is a pretty powerful one as well.

I also noted that my bottom ad does not have a shabby CTR, so if only 7% of my audience is seeing it and it still has a healthy CTR, that means that my policy of placing ads at expected exit points is a good one.

So, while I am not going to be able to use this information for my original plan, I am going to keep tracking this number and trying to improve it. Seems to me that providing information that someone actually finishes reading is a better indicator of quality than time on page or bounce rate.