Removing Adsense From Non-Performing Pages

Best advice I've gotten to date

[spaceylacie]

I've been listening to advice on this board for about a month now and have nearly tripled my eCPM in that time.

The best advice I have received is to remove Adsense from my pages that weren't performing. This somehow made everything go up, throughout entire sites.

Channeling and testing out different border colors and formats was the second best advice, second because I had already sort of been doing this.

Thanks y'all!

[ddd5280]

Do you have any ideas why taking off the ads of poor performing pages works well for you?

[alika]

Smart pricing reasons - you remove the non performing pages and are left with pages that generates good CTR and possibly conversion. Then G rewards you with higher EPC.

The more you keep non-performing pages around, then the more it drags you down. Hence you'll get lower EPCs.

[Sally Stitts]

Is there a "rule of thumb", for dropping the low-performing pages?

For instance, simply remove the lowest 10% of poor performing pages? 5%? 20%?

Or is the gap significant, between the best-performing pages and worst-performing pages? Should one remove pages that perform a given percentage less than the top pages?

Or is it an absolute - remove all pages below a given CTR, no matter what?

Any strong opinions on this? I think that we would all like to know the best course of action.

[gamb]

is this low performing pages on a specific site, or low performing pages on any site you use adsense on?

[spaceylacie]

I have no idea why this worked on my site, the smart pricing I suppose. I just know that it worked.

I removed Adsense from positions that were performing very low compared to other pages. Such as a .5% CTR compared to a 3-5% CTR. So, everything making about 10-25% of what other pages were doing got their Adsense removed.

It worked! EPC is up about 25% or so.

[Sally Stitts]

Too late, sorry.
If you don't respond quickly enough, your post can look stupid. I guess there is no way to just remove the post.

[edited by: Sally_Stitts at 3:47 pm (utc) on May 20, 2005]

[spaceylacie]

In answer to Gamb, the lowest performing pages on each site, compared to the other pages on that site only.

I saw a big gap between my good performing pages and the lower ones, so it was easy to know which to remove Adsense from. Looking back at which I removed, I'd say everything under 20-25% of the other pages on that site.

[jschmitz]

Is this site specific? For instance, I have site that performs really well using adsense, another where the adsense just doesnt work. Does my 'bad' site pull down my numbers overall, or are mynumbers contained within the site stats?
tks

[oddsod]

>> Does my 'bad' site pull down my numbers overall

That doesn't seem to be the case from my experiments.

[jschmitz]

whoops, sorry, saw new posts, got answers. you guys are too fast for me!

j

[spaceylacie]

They seemed to be contained within site stats. I did it with one site at a time, EPC for just that site went up(it took a couple days), other sites were uneffected. Same with other sites, noticed changes to that site only. Stats from one site don't seem to effect other sites.

I have one "bad" site too, I don't think it's negatively effecting my other, better performing sites.

[hunderdown]

Good to see others posting about this -- I suggested that dropping low-performing pages could boost earnings by boosting EPC several months ago, and was met with considerable skepticism. But it looks like enough people have seen it happen that it's a real effect.

[spaceylacie]

I read about it on here and people were disagreeing with the person who said it at the time(maybe it was your post, hunderdown, I don't remember).

It made a lot of sense and the person who said it seemed to know what they were talking about. So I tried it. And it worked!

[spaceylacie]

P.S. I should have stressed more that it takes a couple of days for the changes to take place, even on a site that googlebot visits often. You could even see a lower EPC(but should stay about the same) for the first few days before the algo kicks in, or however that works. Once it kicked in for me, the higher EPCs seem to be here to stay.

[jomaxx]

This is not my experience. I recently added AdSense to a large number of pages where the CPM is somewhat marginal, and I actually saw the CPM go up for the whole site.

That's just one data point, of course. The most important question is probably what "low-performing" means. Low CTR and low CPC are normal situations and IMO not necessarily bad. Poor targeting or an utter lack of relevant ads is another matter.

[spaceylacie]

All of my pages, in general, show very targetted ads, so this wasn't a factor.

By low performing, I mean low performing, look at this page:
19,024 12 0.1% $0.05 $0.99

Over 19,000 people, 12 clicks, and just under a buck. Compared to other pages with eCPMs over $5.00. You'll know when you see a low performing page!

[europeforvisitors]

Smart pricing reasons - you remove the non performing pages and are left with pages that generates good CTR and possibly conversion. Then G rewards you with higher EPC.

That hasn't been my experience. My image-gallery pages have an extremely low CTR, which results in an abysmal eCPM, but EPC for those pages is considerably higher than for my site as a whole.

The original post in this thread was talking about eCPM, and it's important to remember that eCPM is determined by EPC and clickthrough rate. If you have pages with low CTR and you remove their AdSense code, your eCPM will go up simply because your total earnings are being divided by a smaller number of impressions for reporting purposes. That has nothing to do with "smart pricing." Removing AdSense code from those low-CTR pages may give you a higher eCPM, but it will also lower your total revenues.

Unless you have clear evidence that your EPC is being harmed by ads on specific pages or directories, you should remove the AdSense code only if you can replace the ads with higher-paying alternatives.

[spaceylacie]

Since you have higher EPC on those pages, it might not work in your case. My low pages were low everything, low CTR, low EPC, low eCPM, low bottom line.

[elsewhen]

another explaination of this phenomenon (i dont have evidence, i am just theorizing), is that the removal of low paying ads on your site, is making some visitors stay longer, and some of them may hang around to get to a higher paying section of your site, and then click.

so to use your example spacie... those 12 people that clicked on the low-paying ads maybe stuck around on your site, and visited some other section, and clicked on ads that were much higher paying.

in other words, you are getting rid of the low-paying exits of your site, and users poke around, and a certain percentage find higher-paying exits.

this effect would be particularly evident for pages on which adsense is the only site-exit (ie not other outgoing links).

[oddsod]

Er, EPC is an average. So, when removing the lower paying clicks, even if you don't gain EPC on the other clicks your overall EPC will go up. No?

[spaceylacie]

You know, actually, the pages I removed the Adsense from WERE pages where Adsense was the only exit, all the other links on the page lead to other areas of my site. I think you may have hit on something there.

P.S. about a prior comment made, my eCPM went up as a whole, for a variety of factors. One of which was increasing my EPC by dropping Adsense on non-performing pages.

[spaceylacie]

"Er, EPC is an average. So, when removing the lower paying clicks, even if you don't gain EPC on the other clicks your overall EPC will go up. No?"

The pages I removed it from didn't necessarily have lower EPCs, my average EPC is not very high to begin with. As a whole, they were just slightly lower than other pages, but now my EPC is up over 25%. And, it was only a few page that I removed it from compared to hundreds that I didn't.

So averages, just because some were dropped, no, it wouldn't factor in that much.

[spaceylacie]

Oh, and before anyone jumps down my throat over that comment, IN MY CASE, it didn't factor in that much.

A general answer to that question would be, yes.

[hunderdown]

My experience was that overall EARNINGs went up, not just average EPC. If the only positive effect was that taking out lower-paying clicks increased the average EPC, overall earnings should go down.

Possibly this isn't going to work for everyone. It didn't work for EFV. And in my experience it's only worth doing, as spaceylacie said, if the low-performing pages are really low-performing--in my case, the CTR was one-sixth or lower the site average.

Maybe what it comes down to is this: even if the better earnings aren't the result of a site's EPC going up due to better CTR, removing low-performing pages can benefit you. Maybe it's just that it reduces ad blindness if visitors see fewer ads. Maybe it's keeping them on the site so that they can click a better ad, as someone suggested (I'm doubtful about that for my site, though). Doesn't matter why. The point is that can be a worthwhile thing to do....

[europeforvisitors]

Maybe what it comes down to is this: even if the better earnings aren't the result of a site's EPC going up due to better CTR, removing low-performing pages can benefit you...

Or maybe not, once site-targeted CPM ads enter the fray. I can see those being a real income generator for sites that can withstand an advertiser's scrutiny but don't attract high EPCs and/or CTRs on every page.

[spaceylacie]

Yes, perhaps site-targeted CPM ads will be a whole other story once we start seeing them. I'm wondering how that will work, if we'll get a choice of whether or not / where to run them. What kind of overall effect it'll have...

[Juan_G]

hunderdown wrote:

Good to see others posting about this -- I suggested that dropping low-performing pages could boost earnings by boosting EPC several months ago, and was met with considerable skepticism. But it looks like enough people have seen it happen that it's a real effect.

Probably you are referring to this thread:

Adsense on non performing pages
[webmasterworld.com...]

Some recent threads on this controversial -and interesting- topic:

Less pages with ads, higher earnings?
[webmasterworld.com...]

Drop in CTR + Rise in EPC
[webmasterworld.com...]

[Juan_G]

And another interesting one related to the less adds higher earnings topic, in the ad blindness ramification:

"Rotating" AdSense ads with images test
[webmasterworld.com...]

Surely we should also keep in mind EFV's good point on the new site-targeted CPM, that wasn't discussed yet on those previous threads.

[badtigger]

isnt this all just as simple as this:

Say you have a site with 400 pages, but 50 of them are not getting significant CTR. So, you remove them, and the CTR rate would necessarily go up, as an approximate number of clicks across the whole site are coming from far fewer pages.

So, the end result is that this increases your earnings, becuase there are more targeted clickthoughs.