Hiding CPM ads in the footer where most users won't see them doesn't sound like a very good long-term strategy for keeping advertisers.

i've read that putting ads at the bottom of the page works on some websites... typically at the end of an article, for instance.

remember that there are contextual image ads as well as site-targeted ads... fwiw, you could set up a site-targeted channel at the bottom of the page, but it won't stop site-targeted ads from appearing in other ad blocks on your site.

i've read that putting ads at the bottom of the page works on some websites... typically at the end of an article, for instance.

Sure, and it's fine for a publisher to try (or risk) putting a CPC ad unit at the bottom of the page if he or she wants to. But CPM advertisers want placement "above the fold" because they're paying by the impression. If an ad is "below the fold," advertisers are likely to be paying for impressions they aren't actually getting (since not all users will scroll down to where the ads are displayed).

We find that contextual adverts at the end of searched for data perform well. It is logical. The suer has deliberately searched for some data, read it, and gets to the end only to find related links.

Matt

We find that contextual adverts at the end of searched for data perform well. It is logical. The suer has deliberately searched for some data, read it, and gets to the end only to find related links.

Sure, and I wouldn't disagreee, but the OP was talking about CPM ads, which are site-targeted (not contextual).

the o.p. never claimed that his footer block was beneath the fold... what if his pages are short? the average screen size is 1280x these days, so 800x600 sites are not realistic anymore.

site-targeted ads don't function like direct advertising on a site does, because very few adsense advertisers are going to hassle with looking for specific cpm placements on specific sites... that's one reason they pay so poorly, google does not give publishers the tools to be specific on which ad blocks should never have any site-targeted ads.

so you can put an ad block on the bottom of the page, and be comfortable with the fact that it probably won't make the slightest bit of difference to the advertiser, because they'll never know.

Well, he used the phrase "even if no one is clicking on it," which implies that he doesn't expect the ad unit in the footer to draw much (if any) attention.

As for your comment that "they'll never know," we'll just have to disagree on ethics and fairness to advertisers. And by the way, not all advertisers are as ignorant as some publishers apparently think they are. See the AdWords Forum thread on site-targeted banner fraud at:

[webmasterworld.com...]

site-targeted ads don't have good ctr to begin with, so his expectations are realistic.

if the advertiser is getting his moneys worth, your cute little "ethics and fairness" dig is not relevant.

and if the advertiser is too lazy to care where his ads are placed, it's usually because he's not paying much to begin with... that is the typical site-targeted situation.

>>>See the AdWords Forum thread on site-targeted banner fraud<<<

perhaps you could just tell us how fraud is relevant to putting the banner at the bottom of the page?

Look, the guy wanted to know what other people thought. I gave him one point of view, you've given him another. He'll have to make up his own mind. End of story.

I don't know about sticking em at the bottom of the page, but I do know from experience that if your site is targeted you should have the maximum amount of ad units you can. This is so you can display the banners AND the contextual units at the same time.

I would place them "all" in good locations because you never really can tell "which" ad unit will display a banner and which will display a contextual unit.

I don't know about sticking em at the bottom of the page, but I do know from experience that if your site is targeted you should have the maximum amount of ad units you can. This is so you can display the banners AND the contextual units at the same time.

There's an interesting subtext to this. And it's a bit complex. On the surface, the above makes perfect sense. But it may not -- it depends on advertiser behavior which we don't know. And the changes in progress that allow advertisers to choose a specific ad block also come into play (or will, it's not "there yet")

You are assuming (and it's a fair assumption) that advertisers don't really look at the sites they are advertising on, and given the horribleness of the advertising process, no doubt a lot of advertisers aren't looking too hard, or don't care, choosing to look at ROI after the fact.

But I'll also say there are some advertisers who look at prospective sites more carefully and do care. We're one. I've probably looked at 100-150 sites with an eye to site targeting them, and in almost ALL cases we decided not to because of

1) too many adblocks on each page which dilute our ad investment.
2) terrible placement of the adblocks which almost guarantee we'll lose money by advertising there because nobody will see the darn things in the clutter, or because they are at the bottom.

So, if you think advertisers really aren't bothering to look or don't care about the quality of the ad spaces, then it makes sense to have a lot of blocks.

If you believe they ARE caring (or will care) and want to attract higher rates, then fewer blocks well positioned (so called premium blocks) would be the obvious choice.

For site targeted ads, we look to place on (obviously) highly relevant sites, in a prominent position, with no visual clutter, so we will get max targeted clicks. I'm willing to pay serious bucks IF I can find such sites (no great luck so far), but multiple ad blocks kills any deal NO MATTER HOW RELEVANT THE SITE IS.

Google is trying to address some of these issues with new features and programs. It's WAY far off.

adsense was never, and still is not, designed for site-targeted cpm... maybe someday google will give publishers and advertisers the control to communicate about things like the premium placements that efv was referring to.

until that happens, most sites will continue to earn a lot more money with contextual; that is the only site design priority that they need to have, and contextual typically gives a better page ecpm with more than one ad block... in some cases it can even be a phenomenally high page ecpm that you could never achieve with static banner-based advertising.

advertisers who don't want to pay for contextual should do everyone a favor, and just advertise directly with the site itself... that's the only kind of web advertising that i do these days.

advertisers who don't want to pay for contextual should do everyone a favor, and just advertise directly with the site itself.

It isn't that simple, because AdWords/AdSense advertisers who use site targeting don't necessarily want to enter into contracts with a slew of different sites.

Time may be of the essence, too. For example, an airline that wants to promote a last-minute fare sale (that's a real-life example, by the way) may find it easier to buy a multi-site flight of site-targeted CPM ads from AdWords on short notice than to go through the usual process of having its agency design and place display ads.

Still, there's no doubt that, for the most part, site-targeted AdSense ads are a side show, not the main event. And that's just what they were designed to be: a way to improve income for Google and publishers at moments in time when higher-paying contextual ads aren't available for specific ad units and keyphrases.

1) too many adblocks on each page which dilute our ad investment.

Not to sound cold hearted, but that's just too bad. My job is to maximix revenue for my site while preserving it's intergrity. If my pages are large enough to support the maximum amount of ads, then so be it.

2) terrible placement of the adblocks which almost guarantee we'll lose money by advertising there because nobody will see the darn things in the clutter, or because they are at the bottom.

I agree 100% I am not talking about terrible placement. I am talking about optimal placement, achieved by experimentation over time. Terrible placement will not give you maximum CTR. If your running max amount of ads, and you don't know which unit displays the image and which doesn't in just makes good common sense to have them all placed correctly.

Here's an update on my post:

I've done a quick test on my page today. I placed an image ad at the footer, but it is still visible without scrolling (yes my page is short)

Result is disappointing, over 9000 impressions, I got twenty cents only....
I tried replacing with text ad at the footer but it looks too ugly. So I just remove the footer ad completely.

My page performed very good with site-targeted ads a week ago. Now those ads are gone and my earnings drop by 25%. I'm trying to catch up with the difference so I'm seeking ways to optimize the earning potential.

Not to sound cold hearted, but that's just too bad. My job is to maximix revenue for my site while preserving it's intergrity. If my pages are large enough to support the maximum amount of ads, then so be it.

I bet you started responding before you read my entire post. No argument with the above. My point was that if you want or hope to attract decent paying cpm ads, consider that certain kinds of placements may push potential advertisers to go elsewhere. I won't repeat what I said in the message, but that's the main point.

>>>I placed an image ad at the footer<<<

as i mentioned earlier, just because it's an image ad block doesn't mean that it's a site-targeted ad channel.

as i understand it, what you want there is a site-targeted ad channel, something that the advertiser can specifically choose... afaik, you can't do that with a contextual image ad.