|What makes a "low performing page" a low performing page?|
Is "low performance" subject to a "tune up"? New plugs? New wires?
| 12:01 am on Jun 13, 2006 (gmt 0)|
I've seen multiple references to "low performing" pages and suggestions that removing Adsense from "low performing pages" (LPP) was a good approach.
Removing a "low performing page" seems a bit draconian. Pity the poor page. Cut down in its prime. A cast off. So sad. :(
Why not give a LPP an examination and some vitamins? An infusion? A bit of cosmetic surgery? A verbiage tune-up?
Before you remove a LPP do you page-cullers first attempt to determine what may be making the page perform in a lowly manner?
When you identify a LPP - whick I'll take to be a page with traffic but without clicks - do you look for ways to improve on that?
Any tips of general application for general maintenance of low performing pages?
Have you learned anything about performance by looking at LPPs? What is revealed by LPPs that might be useful in crafting HPPs? (Ya, high performing pages.)
Can you take what you learn from studying LPPs and transform that into knowledge about how to craft a HPP? Any insights to share?
| 2:33 am on Jun 13, 2006 (gmt 0)|
One difference on my site is that two particularly low-performing pages have information I update regularly. They are magnets for regulars, who check the updates and disappear. Content is also rather unfocused but I think the main problem is the way that people use the page.
My point is that "low-performing" isn't just determined by on-page factors....
| 7:02 am on Jun 13, 2006 (gmt 0)|
Yes, that's a problem of many -not all- news and community sites, but some topics tend to be lower performers than others.
Also, general pages are often low-performing. Pages focused on a specific topic are better performers.
| 7:38 am on Jun 13, 2006 (gmt 0)|
One reason for the low performance could be the keyword. Perhaps the keyword is low performing, brining in a low number of visitors or the keyword is of such a nature that the visitors do not click on the ads for this particular keyword.
You could, of cause, consider changing the keyword to something more yielding or changing it to something that is more click motivation. Maybe if you collect a number of keywords from low performing pages, you would be able to identify a pattern as to why they do not perform
| 8:27 am on Jun 13, 2006 (gmt 0)|
Too much tampering. Too many changes. Fiddling with the page by changing keywords continuously etc, generally brings down EPC for me. Once I am happy with a page, I leave it for a while and things begin to stabilize and then move up slowly.
I will say too much experimentation, creates a LPP.
and Yes .. More focussed the page, better the ad targetting and better EPC.
| 2:06 pm on Jun 13, 2006 (gmt 0)|
So, low performing page means one or more or the following situations likely exist, which are as much connected to the ad inventory as the page's content and mark-up:
- Poorly targeted ads, ads that are off topic for page [Solution: Section targeting, tighten page topic/focus]
- Available ad inventory for page topic lacks effective "call to action"
- There is scant inventory for the page topic, so you may be getting PSAs or more remotely related ads
IF the nexus of the problem of a LPP lies with the ad inventory then it seems amiss that a publisher should feel they have to take steps to remove ads for fear of some form of revenue penalization.
IF the nexus of the problem lies with G's misreading of the gist of the page then it seems amiss that a publisher should feel they have to take steps to remove ads for fear of some form of revenue penalization.
IF the nexus of the problem is something about the page itself - which includes visitors who are drawn to the page by deceptive methods - then I can see a basis for reducing payments based upon a considerable volume of ad delivery services without a) clicks; or, 2) conversions.
I would think, on the whole, that removing LPPs would be counterproductive - assuming one knows that they have crafted an otherwise well structured and topically targeted page.
Thanks for the insights about low performing pages and what to do.
Feel free to continue beating me around the head if I've failed to grasp something important. :)
| 2:12 pm on Jun 13, 2006 (gmt 0)|
Why would removing ads that aren't performing from an otherwise popular page be counter-productive? I can leave the page ad-free, thereby improving the look of my site and reducing ad blindness if people navigate elsewhere. Or I can substitute a different ad.
I agree that IF the traffic is targeted and the page is focused, then even if the CTR is low it's worth leaving ads on, because conversions will be good. But if the traffic is coming for specific information and leaving, not interested in the ads, why show them ads?
| 2:20 pm on Jun 13, 2006 (gmt 0)|
Most excellent point - "it's the visitor experience stupid" :) - and a well deserved knock on the noggin.
Well, I did ask for it, didn't I? :)
P.S. I'm not saying you called me a knucklehead. I'm calling myself one . . 'cause on this point I was.
| 10:22 pm on Jun 13, 2006 (gmt 0)|
I'm still not clear what people mean by a low performing page. Is it a page with low CTR? Or is it low EPC? Or is it a page that just doesn't get many visitors so doesn't make much money?
My impression is that only low EPC pages might hurt our overall EPC. Or does low CTR or traffic hurt too?
| 10:38 pm on Jun 13, 2006 (gmt 0)|
|My impression is that only low EPC pages might hurt our overall EPC. Or does low CTR or traffic hurt too? |
In our experience, it's the low eCPM that tends to drag down the whole site. It's not necessarily the fault of the page or even the inventory of advertisers, it's what they are willing to pay and whether your visitors click. You could have the most popular page on the web about, say, inexpensive ball point pens, and there may be plenty of office supply outfits and gold and jewel encrusted ball point pen sellers bidding high on the keywords, but their click through could be near zero, ergo lousy eCPM. There may also be some genuine cheap ball point pen sellers, but how much can they bid on an ad? Even if their click through is good, you'll still have a low eCPM.
Whatever caused the page to be a low earner, in the experience of many of the publishers on this forum, you'll earn more in the long run if you remove Adsense from it. It's kind of surpising that the Adsense algo can't figure all this out and keep the earnings maximized for everybody, but it must be following some rules that get in each others way. My guess is that like many code monsters, the creators aren't sure exactly how it functions anymore, they just try to tweak it for the best overall results.
| 10:54 pm on Jun 13, 2006 (gmt 0)|
I did post that I remove ads from LPP but that does not mean they do not make me money, I still use other networks, also referrals.
LPP pull their own weigh and make 1/3 of my site's earnings.
I want to add another possible reason for LPP to your list, which is having non supported languages.
| 11:38 pm on Jun 13, 2006 (gmt 0)|
|I'm still not clear what people mean by a low performing page. |
There are different methods to identify low-performing pages:
- Low EPC comparing with AdWords rates for the page keywords. This is a very interesting method to find smart-priced pages after low conversions for advertisers, but I've not tested it yet because it requires quite a lot of time and research. Apart from this it seems an excellent method, explained by 21_blue on the library thread How to make Smartpricing work to your advantage [webmasterworld.com].
- Low eCPM. A faster method (with no AdWords research), probably not so reliable like the previous one, but it also seems to work reasonably well, according to my testing. I explained some points on that linked thread.
- Low CTR. A controversial method, also fast, and perhaps used due to the current limit of 200 active AdSense channels, a low limit that leads many publishers to use alternative trackers. Those trackers use indirect methods to analize CTR (more or less reliably), but of course they do not have access to EPC and eCPM data beyond the AdSense channels.
So, if you have time for research, I would suggest using the EPC/AdWords method with your 200 most visited pages. On the other hand, if you are busy, you may try the eCPM method instead for those 200 pages (or channels), and maybe a test on the CTR theory for other less visited pages.
A possible smart pricing effect from removing low-performing ads, or at least the beginning of an effect, should be visible in some days, or one or two weeks.
In any case, with or without smart pricing benefit, just to improve the user experience is enough reason to remove those ads that are not being useful to our readers.