CPM much lower for asynchronous ads

CMP, Asynchronous, synchronous, adsense, performance

[guggi2000]

In one specific part of our website we have monitored the performance of synchronous vs asynchronous Adsense ads. We have measured over 3 million pageviews and the ad-code was hardcoded, no DFP server.

The synchronous Adsense performed 60% better.

Did anyone else have the same experience?

[Sally Stitts]

Nice to hear this.
I have not gone asynchronous for several reasons.
The biggest reason is that my web editor obliterates Google's asynchronous AdSense code.
Synchronous code works just fine.
Looks like I have stumbled upon the best approach, in spite of myself (Ha!), according to guggi2000.

[nomis5]

Were they mobile, tablet or desktop views. If a mixture what were the %s?

[Rasputin]

That's very interesting, thanks for sharing.

I just today started testing the same thing but mostly because I wondered if the delay in loading the top ad in the articles was causing visitors to have moved on down the page before the ad turned up (the pages are long with lots of photos and I have added lazy load at the same time).

Presumably this was the only varying factor in your test - eg all were responsive or all were non- responsive?

[surfgatinho]

I remember reading this somewhere before. Makes sense given the fact asynchronous ads sometimes don't even render until after you've read past the fold.

[guggi2000]

@Rasputin Yes, only the code changed. Same position, size, layout...

@Nomis5 Desktop 50% difference and Mobile 70% difference

[robzach]

Got 35% better CPM and Earning after switching to sync code.
However, anyone know how to make this ad code responsive like the async code?

[Rasputin]

This is an important problem but I think the answer is no, and thorough reading of the terms seems to suggest it is not permitted just to delete the 'async' part of the responsive adsense code.

If my sync - async test shows positive results I am planning to try using a php mobile detect script and then show the appropriate adsense code based on that, which I think will work albeit less accurately than a proper responsive solution.

That would also allow link units to be shown on larger screens (link units are not yet available in responsive form)

I'd love to here if anyone has found a better solution, or reasons why this won't work...

[MikeNoLastName]

Noticed this some time ago too. The problem, at least in our case, seems to be that the ads near the top, in asynch mode take SOOO long to load that they don't even come up before the user scrolls past them! Until G fixes this long delay, we've made ALL ads above the fold, and in many cases on the first scrolled page synch and let any others lower than that asynch to help with overall-load time. The load speed is the highest concern or we'd can asynch ads altogether, but our server access and page sizes are such that they are normally very fast anyway, so overall even with the synch ads aren't too bad.
The only problem is now with G using mobile-friendliness as an algo factor and responsive ads are only available in asynch, we're forced to make all above-the-fold ads mobile-size/synch even for desk-tops. This means usually a 300x250 or 320x100 only

[webcentric]

I agree that you probably shouldn't try to remove async from responsive ads. Also, I wonder if other async code on the page could be playing a role in the above results e.g. Google Analytics script loading before Adsense scripts. Anyone know more about this? I'm not a wizard at prioritizing multiple async scripts and not even sure if it's possible (other than the order they appear in the markup). Guess I'm questioning if all the pieces of this puzzle are on the table yet.

[netmeg]

My dev is trying to figure out ways to improve performance in WordPress and with all the various technologies we use (including ads) by using Google Tag Manager - I wish I could tell you more specifically, but it's *whoosh* over my head technically speaking. Once he gets it all finished, I'll make him write up something I can post.

[webcentric]

@netmeg I think the CMS in play has something to do with this discussion. Many here are using WordPress and it has it's own special set of bandwidth issues. I don't use it (build my own CMS sites from scratch in .Net) and I'm able to keep a lot of these issues at bay. I don't have the volume of visits that would support a definitive conclusion either way in this discussion so I'd like to ask the OP

@guggi2000 are you using WordPress? If not, what are you using for a CMS? I think it's a relevant factor in this discussion.

[guggi2000]

@webcentric No CMS is used. We write the code.

Also, I wonder if other async code on the page could be playing a role in the above results e.g. Google Analytics script loading before Adsense scripts.

Very good point. Anyone done some research?

[netmeg]

(I'm using GTM to load Analytics as well)

[breeks]

My media dot net ads were hanging up load time on my sites. Both wordpress and html sites. Contacted them one day later they had new code for me, problem solved. Now ads are lightning fast.

Don;t make much with media only about 25.00 a day but support was excellent.

If only Adsense could do the same.

[super70s]

I've had a Media Net guy after me pretty heavily to use their ads after I signed up last October, I even agreed to a live phone conversation with him last week.

I just don't like that "topic" approach of theirs, I can't see those ads getting any clicks.

[jetteroheller]

Changed April 21nd to async
Compared to the 3 weeks before, not about 7% higher CPM

[guggi2000]

@jetteroheller you have to do an A/B test where one half of your ads are shown with Synch Code and the other half with Asynch. It needs to be at the same time and the same placement position. Give it a least a few weeks.
Any other method has to many outside factors...