Welcome to WebmasterWorld Guest from 34.236.145.124

Forum Moderators: martinibuster

Message Too Old, No Replies

Unperforming Ads & Pages

To improve eCPM...

     
3:38 am on Jun 18, 2006 (gmt 0)

New User

10+ Year Member

joined:Feb 21, 2006
posts:29
votes: 0


Hello everyone.

Regarding the much discussed, "removing unperforming ads, or pages, to improve overall earnings per click".

I want to verify one thing.

If I have a site, with many pages...

And some of those pages get few visits, and NO clicks to speak of, but are still displaying ads...

Will I be able to increase the earnings on my pages which have ads which DO get clicks by removing Adsense from the previously mentioned non-performing pages?

Thanks for verifying this for me, just making sure I have it straight in my head!

5:30 am on June 18, 2006 (gmt 0)

Preferred Member

10+ Year Member

joined:Jan 15, 2004
posts:612
votes: 0


saikyo,

I would not remove Adsense code from pages that get few visitors. Such pages still have the potential of getting a decent CTR.

Just consider a page with 49 page views this month and no clicks. "Lousy!" you say and consider removing the ad. While you think about other ways to monetize, one of your visitors hits the page (page view # 50) and clicks the ad! All of a sudden, that underperforming page (0% CTR) has been turned into a well-performing page (2% CTR).

You should focus on pages that do get decent traffic but no or few clicks, i.e. if you have served 499 pages with no clicks, even a click at pageview # 500 will not turn the page into a real "performer" (CTR 0.2%).

5:58 am on June 18, 2006 (gmt 0)

Senior Member

WebmasterWorld Senior Member 10+ Year Member

joined:June 10, 2004
posts:2076
votes: 0


If you have lots, and lots and lots of poor performing pages maybe, but I'd be inclined to let pages with low numbers of visitors stay. The pages that I remove are ones that have a high number of visitors and no clicks. They have the potential to seriously screw up epc.

A page that has low visitor numbers might in fact actually pay quite well in terms of eCPM, and a load of low value pages that pay quite well considering visitor numbers adds up to a good return overall.

It may be that it's not the page that has low clicks, but certain banners on it. For example if you have several banners on the page then monitor the performance on all of them to make sure they work for you. I think that having banners on what seem to be low value pages on the face of it is not a problem - putting loads of banners on pages that do work is.

7:41 am on June 18, 2006 (gmt 0)

Junior Member

10+ Year Member

joined:May 25, 2006
posts:144
votes: 0


Put another ad service on low performance pages. It will keep your EPC high on Adsense and you can still have the chance of earning money on those unpopular pages.
9:42 am on June 18, 2006 (gmt 0)

Preferred Member

10+ Year Member

joined:May 30, 2005
posts:456
votes: 0


This is yet another AdSense myth (in other words, nonsense).

In regard to Smart Pricing, "low performance" (i.e. low CTR, eCPM or EPC) of ad units is utterly irrelevant. It is easy to show that the previous statement is true:

Smart pricing is based on a quantity that is unknown to publishers: Conversion (that is how many people actually purchase the advertised product or services, etc). As you don't know how well the ads on a particular page convert for the advertisers, you can't determine if an ad unit is "low performing" in regard to Smart Pricing.

If you still don't understand, look at this example:
Page 1: CTR 5%, eCPM $20
Page 2: CTR 1%, eCPM $5

You can see that Page 2 is "low performing" (in terms of CTR and eCPM, when compared to Page 1). Some people here would now advise you to remove ads from Page 2 to avoid Smart Pricing. BUT, they don't know that conversion ratio of Page 2 is 100%, while Page 1 converts only 0.01% [examples]. So if you are afraid of Smart Pricing, in this particular case you should actually remove ads from Page 1. The problem is that you don't know the conversion ratio of any of the pages, so removing the ads from any of the pages is baseless and may even be risky.

10:02 am on June 18, 2006 (gmt 0)

Senior Member

WebmasterWorld Senior Member 10+ Year Member

joined:June 10, 2004
posts:2076
votes: 0


Actually, I wouldn't necessarily give that advice given those statistics.

I don't think there is much debate any longer of the effect of removing poor performing blocks - the debate seems to be what constitutes poor performing.

When making the decision, I look at ALL metrics, and Ileave ads in place for a while before I make the decision. Sometimes a page comes goo, or produces a good average over time, sometimes they simply never work with adsense, and those are the blocks I'm looking to dump.

I look at eCPM, page views,impressions, clicks, ctr, history - the lot. The only advice I can give is to be familiar with patterns on your site and know over time what could be regarded as normal. By doing this you get a better feel for what's going on and what ads really don't work, and are in a better position to make the right decision.

10:08 am on June 18, 2006 (gmt 0)

Preferred Member

10+ Year Member

joined:May 30, 2005
posts:456
votes: 0


the debate seems to be what constitutes poor performing.

As a publisher you have access only to the following figures: eCPM, average EPC, CTR, and number of impressions. Hence, a publisher can call a page "low performing" based only on these figures and on nothing else. None of these figures affect Smart Pricing. Thus, removing "low performing" (in terms of eCPM, CTR, etc.) pages to fight Smart Pricing is absolutely baseless (and is a bad piece of advice -- unfortunately, quite frequent here).

[edited by: John_Carpenter at 10:12 am (utc) on June 18, 2006]

10:10 am on June 18, 2006 (gmt 0)

Preferred Member

10+ Year Member

joined:Jan 15, 2004
posts:612
votes: 0


John,

I think you are right about the Smartpricing here. However, the OP mentioned that "some of those pages get few visits, and NO clicks to speak of", so he was not really referring to Smartpricing. I guess he was seeking advice on whether to remove ads from these pages or not.

I recommend to not remove the ads as
(1) the sample size is too small to make any valid statements of the potential performance. E.g. a page with zero clicks may turn into 2% CTR by just one single click.
(2) a prediction of the performance (i.e. sales conversions) can not be made at all as we do not have insights into this matter anyway.

10:18 am on June 18, 2006 (gmt 0)

Preferred Member

10+ Year Member

joined:May 30, 2005
posts:456
votes: 0


so he was not really referring to Smartpricing

Actually, all these pieces of advice about removing "low performing" ads are inherently about Smart Pricing (the Smart Pricing element in those threads is implied). Why? Because, only Smart Pricing (and nothing else) could increase your overall earnings afer you remove ads from selected pages.

[edited by: John_Carpenter at 10:21 am (utc) on June 18, 2006]

10:19 am on June 18, 2006 (gmt 0)

Preferred Member

10+ Year Member

joined:Jan 15, 2004
posts:612
votes: 0


Point taken. :-)
12:49 pm on June 18, 2006 (gmt 0)

New User

10+ Year Member

joined:Feb 21, 2006
posts:29
votes: 0


Actually, all these pieces of advice about removing "low performing" ads are inherently about Smart Pricing (the Smart Pricing element in those threads is implied). Why? Because, only Smart Pricing (and nothing else) could increase your overall earnings afer you remove ads from selected pages.

Geeeesh, I step out for dinner for just a sec, come back, and there's a war... well, a heated discussion going on!

From this, does this mean, that if I remove said ads from the pages with low visits, and low click through, I could (via smart pricing) increase the earnings I get from my pages that ARE performing.

Is this right?

1:07 pm on June 18, 2006 (gmt 0)

Preferred Member

10+ Year Member

joined:Jan 15, 2004
posts:612
votes: 0


From this, does this mean, that if I remove said ads from the pages with low visits, and low click through, I could (via smart pricing) increase the earnings I get from my pages that ARE performing.

There are two effects that may overlay each other.

1) Earnings-per-click may be linked to CTR. Poor performing pages (in terms of CTR) drag down the overall "valuation" of your site in the eyes of Google, i.e. if you have pages that get eyeballs but where the ads are not clicked, you will earn less per click than on a page with high CTR.

2) The conversion of the visitor into real sales on the advertisers page is said to be linked to the overall payout of your site. This effect is also known as "Smartpricing". If advertisers report that visitors coming from your site are actually buying stuff (or performing another measurable action), then you are a valuable publisher in Googles eyes. Well, in reality its vice versa - if your site does not perform, then you get "smartpriced", i.e. the average EPC will be lowered to reflect the poor ability to convert visitors into sales.

Now, as you can see, EPC seems to be linked to both concepts (CTR on your site, and conversion on the advertisers site).

If you have a lot of pages with FEW VISITS, then you will want to keep the ads on these pages. A single click can turn a 0% CTR into, say, 2% CTR.

If, on the other hand, you are having pages with lots of visitors that do not click the ads, then you may be able to increase your sites value by removing the ads. As the number of clicks remains the same with reduced page impressions, the CTR goes up, and EPC =may= go up as well.

Removing ads will, however, not influence the ability to convert a visitor into a buyer! You do not know whether that single click on a poor performing page will actually convert or not.

To summarize, if you are just talking about low-traffic pages, I would recommend to keep the ads. If you have a high traffic page with low clicks, then you will want to consider removing the ads.

1:21 pm on June 18, 2006 (gmt 0)

New User

10+ Year Member

joined:Feb 21, 2006
posts:29
votes: 0


Understood!

Thanks so much guys!

My particular case is mostly pages with few visits AND few clicks, so I'll be keeping the ads.

Thanks!

2:57 pm on June 18, 2006 (gmt 0)

Senior Member

WebmasterWorld Senior Member 10+ Year Member

joined:Jan 12, 2006
posts:1309
votes: 0


Smart pricing is based on a quantity that is unknown to publishers: Conversion (that is how many people actually purchase the advertised product or services, etc). As you don't know how well the ads on a particular page convert for the advertisers, you can't determine if an ad unit is "low performing" in regard to Smart Pricing.

I see your point John, but in my opinion -- this same quantity is also unknown to Google. There is no reporting mechanism for Adwords advertisers to tell Google: "Yes... when a user on Site A clicked a particular ad, that user navigated to my Buy page and purchased a fishing rod."

So in essence, Smart Pricing has to be based on the same factors you claim it isn't (i.e. CTR, eCPM, EPC, etc.)

As a publisher you have access only to the following figures: eCPM, average EPC, CTR, and number of impressions. Hence, a publisher can call a page "low performing" based only on these figures and on nothing else. None of these figures affect Smart Pricing.

Like I mentioned above, Google only has access to the same figures, so how could the concept of Smart Pricing be baseless? Google does not track a user (through an affiliate ID or any other measure) after that user clicks an ad. Therefore, Google has no idea whether that user produces an ROI for the advertiser.

Per Google's statement to Adwords advertisers:

We are constantly analyzing data across our network, and if our data shows that a click is less likely to turn into business results (e.g. online sale, registration, phone call, newsletter sign-up), we may reduce the price you pay for that click.

This means that Google is predicting and making assumptions based on its own analysis.

More of the statement from Google to advertisers:

We take into account many factors such as what keywords or concepts triggered the ad, as well as the type of site on which the ad was served. For example, a click on an ad for digital cameras on a web page about photography tips may be worth less than a click on the same ad appearing next to a review of digital cameras.

So, the guy who has a site about photography will get a lower EPC than the guy who has a site about digital camera reviews. Google bases this reduction in EPC (Smart Pricing) on their assessment that the photography site's user isn't really as valuable as the digital cameras site's user.

Again, from Google's message to advertisers:

Google saves you time and hassle by estimating the value of clicks and adjusting prices on an ongoing basis.

The term "estimating" goes back to my original point: Google makes ad adjustments based on what it thinks will work best on your site.

How does this relate to CTR and EPC and the publishers definition of Smart Pricing?

It's all the same because publishers have realized that some pages are well-targeted and it's those pages that grab the highest EPC. Pages that are considered low-performing, are usually those with badly targeted ads.

6:26 am on June 19, 2006 (gmt 0)

Senior Member

joined:May 5, 2005
posts:735
votes: 0


>>>I see your point John, but in my opinion -- this same quantity is also unknown to Google. There is no reporting mechanism for Adwords advertisers to tell Google: "Yes... when a user on Site A clicked a particular ad, that user navigated to my Buy page and purchased a fishing rod."

So in essence, Smart Pricing has to be based on the same factors you claim it isn't (i.e. CTR, eCPM, EPC, etc.)<<<

thank you for that feedback from the adwords side of the fence... i have long suspected that conversion was nothing more than "smart pricing" fud that gets passed around as gospel out here.

if google knows what leads convert to sales, where is the adwords documentation to back that reporting process up?

7:56 am on June 19, 2006 (gmt 0)

Preferred Member

10+ Year Member

joined:May 30, 2005
posts:456
votes: 0


same quantity [conversion] is also unknown to Google. There is no reporting mechanism for Adwords advertisers to tell Google: "Yes... when a user on Site A clicked a particular ad, that user navigated to my Buy page and purchased a fishing rod."

You are wrong. See e.g. [adwords.google.com...]

Thus, the rest of your statements is wrong as well.

[edited by: John_Carpenter at 8:01 am (utc) on June 19, 2006]

7:59 am on June 19, 2006 (gmt 0)

Preferred Member

10+ Year Member

joined:May 30, 2005
posts:456
votes: 0


i have long suspected that conversion was nothing more than "smart pricing" fud that gets passed around as gospel out here.

It's not FUD, but harsh reality.

if google knows what leads convert to sales, where is the adwords documentation to back that reporting process up?

[adwords.google.com...]

8:08 am on June 19, 2006 (gmt 0)

Preferred Member

10+ Year Member

joined:May 30, 2005
posts:456
votes: 0


Earnings-per-click may be linked to CTR. Poor performing pages (in terms of CTR) drag down the overall "valuation" of your site in the eyes of Google

Actually, low CTR (poor targetting) is not a problem of the site but of Google's targetting algorithms and quality of the served ads. Removing ads from low-CTR pages could even be harmful. The low-CTR ads might actually improve the Smart Pricing status of your site if they convert well (see my previous posts).

8:30 am on June 19, 2006 (gmt 0)

Preferred Member

10+ Year Member

joined:Jan 15, 2004
posts:612
votes: 0


John,

I agree, BUT...

low CTR (poor targetting) is not a problem of the site but of Google's targetting algorithms and quality of the served ads

I would like to add that poor CTR can also be caused by the site. For example, I have strong and unique content that keeps users on site. I see a higher CTR on "weak" pages, when people actually have seen/read enough.

(One might argue that I should build more "weak pages", but this will not help getting sustainable traffic. It may boost CTR short-term, but people will then stop recommending the site and thus overall traffic and overall revenue will drop.)

So, we can not simply blame poor targeting if you see a poor CTR. It may be linked to other factors, especially the site hosting the ad, as well.

8:39 am on June 19, 2006 (gmt 0)

New User

10+ Year Member

joined:Aug 30, 2004
posts:39
votes: 0


Theory behind increasing earnings is one thing and there seem to be different opinions about it. Another way is a more practical approach. Trying to improve earnings I removed ads from 'low performing pages' as judged by CTR and eCPM. The result was overall better earnings. So whatever is behind it, what one sees when changing or removing ads from pages, is relevant. Maybe a proper advice would be : if you notice poor performing pages, try to see what happens when removing ads. If it does not work out as you want, just go back to how it was before.
1:02 pm on June 19, 2006 (gmt 0)

Senior Member

WebmasterWorld Senior Member 10+ Year Member

joined:Jan 12, 2006
posts:1309
votes: 0


You are wrong. See e.g. [adwords.google.com...]

Yes...I know about Conversion Tracking, but Conversion tracking is simply a term used to glorify cookie trails.

A conversion cookie tells the advertiser and Google: "Yes, a user clicked on my Adwords ad and has arrived at a page where I think they'll make a purchase." Problem is, not all sites' have ROI driven by PayPal buttons and credit card processing. The ROI for many websites can't be measured by a traditional purchase of a product.

From the Adwords Conversion Tracking info page:

In online advertising, a conversion occurs when a click on your ad leads directly to user behavior you deem valuable, such as a purchase, signup, page view, or lead.

My question is, how would Google know what's "valuable" for one particular website? The algorithm doesn't even properly target ads half the time.

Let's say a site owner decides to take the advice of many by placing a, "Send To A Friend" link on each page. A user who arrives through an Adwords ad and decides to click that link could be considered a "valuable lead"...(since he is browsing the site and decides to tell a few friends about it)

Since the link is included on each page, is each page of your site then qualified as a conversion page?

What if that user never clicks the "Send To A Friend" link and decides to email his/her friend the URL directly, or call them on the phone?

I'm just saying that it's almost impossible for Google to determine what is valuable and what isn't.

Therefore, I believe (like many others) that the Smart Pricing standard is based on targetting, CTR and EPC.

3:04 pm on June 19, 2006 (gmt 0)

Preferred Member

10+ Year Member

joined:May 30, 2005
posts:456
votes: 0


Yes...I know about Conversion Tracking

Really? Then I wonder why you wrote: 'this same quantity [conversion] is also unknown to Google. There is no reporting mechanism for Adwords advertisers to tell Google, [...]'.

A conversion cookie tells the advertiser and Google: "Yes, a user clicked on my Adwords ad and has arrived at a page where I think they'll make a purchase." Problem is, not all sites' have ROI driven by PayPal buttons and credit card processing. The ROI for many websites can't be measured by a traditional purchase of a product.

Yes, not all publisher track conversions. However, that is utterly irrelevant.

My question is, how would Google know what's "valuable" for one particular website?

You don't really know what conversion tracking is (contrary to what you said). Let me explain it to you: As an advertiser you insert your tracking code generated by Google for instance into a "Thank you for your purchase" page. This tracking code reports to Google that the visitor they sent you bought the advertised product.

I'm just saying that it's almost impossible for Google to determine what is valuable and what isn't.

What you are missing is this: It's not Google who determines what's valuable to an advertiser -- the advertiser himself/herself determines (signifies) that (by location of the tracking code).

Therefore, I believe (like many others) that the Smart Pricing standard is based on targetting, CTR and EPC.

Smart Pricing cannot be based on EPC, since EPC is the figure affected by Smart Pricing. And how Smart Pricing could be based on targetting is completely beyond me.
3:06 pm on June 19, 2006 (gmt 0)

Preferred Member

10+ Year Member

joined:May 30, 2005
posts:456
votes: 0


Maybe a proper advice would be : if you notice poor performing pages, try to see what happens when removing ads. If it does not work out as you want, just go back to how it was before

Yes, that would be much better advice.

3:46 pm on June 19, 2006 (gmt 0)

Senior Member

WebmasterWorld Senior Member 10+ Year Member

joined:Jan 12, 2006
posts:1309
votes: 0


Really? Then I wonder why you wrote: 'this same quantity [conversion] is also unknown to Google. There is no reporting mechanism for Adwords advertisers to tell Google, [...]'.

Because a cookie isn't a conversion mechanism. It simply tracks a user across your site. If Google ran a payment gateway similar to PayPal or some other ecommerce entity -- where they actually tracked the number of sales (not some cookie that expires after 30 days) -- then we're talking bonafide conversions. What happens if a user clicks your ad; goes to your site, but doesn't make a purchase for 32 days? Did you just lose your conversion mechanism? Your business gets a customer, but Google doesn't know about it.

Let me explain it to you: As an advertiser you insert your tracking code generated by Google for instance into a "Thank you for your purchase" page. This tracking code reports to Google that the visitor they sent you bought the advertised product.

I already know that, which is why I stated:

"Problem is, not all sites have ROI driven by PayPal buttons and credit card processing. The ROI for many websites can't be measured by a traditional purchase of a product."

From John C:

What you are missing is this: It's not Google who determines what's valuable to an advertiser -- the advertiser himself/herself determines (signifies) that (by location of the tracking code).

Again I ask, will each page of a website be considered a "conversion" page if a user directs other users to the page through a "Tell A Friend" link? Are returning visitors not considered "valuable" since they're not landing on a "Thank You For Buying From Me" page?

From mzanzig's previous post:

...if your site does not perform, then you get "smartpriced", i.e. the average EPC will be lowered to reflect the poor ability to convert visitors into sales.

And how Smart Pricing could be based on targetting is completely beyond me.

Lots of visitors to a particular -- that page has badly targeted ads -- Visitors don't click -- EPC for that page drops -- Smart Priced

11:04 pm on June 19, 2006 (gmt 0)

Senior Member

joined:May 5, 2005
posts:735
votes: 0


>>>Yes, not all publisher track conversions.<<<

publishers can't track adsense conversions, advertisers do that... and if it's accurate data like you claim, how come publishers can't see anonymous conversion data for every page? why can't we use it as a tool for improving our sites?

>>>And how Smart Pricing could be based on targetting is completely beyond me.<<<

that's because you don't know what "smart pricing" really is... and neither does anyone else... like others have already posted in this thread, what we do know is that if you bring untargeted traffic to your pages, the ctr will drop like a rock... which has the potential to wreck the epc across your entire account.

11:22 pm on June 19, 2006 (gmt 0)

Senior Member

joined:May 5, 2005
posts:735
votes: 0


i just did a search on the google adwords groups for "conversion tracking" ...it was interesting, take a look and you'll see how inaccurate the concept is... for instance, look for a post by jeremyc on march 19.
3:27 am on June 20, 2006 (gmt 0)

Senior Member

WebmasterWorld Senior Member 10+ Year Member

joined:July 25, 2005
posts:1010
votes: 5


since we all can only speculate how smart pricing considers which input variables, let's get on a more empirical track.
for a start, to decide which ad blocks to delete, you need a statistically significant amount of traffic on the pages in question.

ctr is likely not a factor in smart pricing algorithm, since there is no need to assume, that one click out of 100 (low ctr) converts worse than one click out of 10 (high ctr).
however, ctr is an indicator for popularity of an ad. if ctr is low, people are not interested. don't annoy your visitors with ads they are obviously not interested in! instead take them to places, where they might pay more attention and the ads are better suitable for their current surfing mood.

epc is also not a factor, but, and this is important, a quite acceptable indicator if smart pricing is at work (as long as the ads on your sites are in a way comparable).

for me, poor performance = low ctr and low epc = kick out the ad blocks in question and recheck your earnings

8:28 am on June 20, 2006 (gmt 0)

Preferred Member

10+ Year Member

joined:May 30, 2005
posts:456
votes: 0


Because a cookie isn't a conversion mechanism

Well, you may criticize the accurateness of their system, but that's how it works and that's what they actually use (no matter if you like it or not).

Lots of visitors to a particular -- that page has badly targeted ads -- Visitors don't click -- EPC for that page drops -- Smart Priced

Again, this is nonsense. Smart Pricing cannot be affected by CTR and this is easy to prove. Example: Ads with CTR of 0.01% could have 100% conversion, while ads with CTR of 50% could have poor conversion of 0.01%. The advertiser will report the low conversion to Google (because the only thing he wants is high conversion, not high CTR + zero conversion) and the ads with high CTR will get your site Smart Priced. I already wrote a similar example and I think it is quite easy to understand if you read carefully and pay attention.

Again I ask, will each page of a website be considered a "conversion" page

Read the Adwords Help. I think I've been wasting my time.
8:36 am on June 20, 2006 (gmt 0)

Preferred Member

10+ Year Member

joined:May 30, 2005
posts:456
votes: 0


>>>Yes, not all publisher track conversions.<<<

publishers can't track adsense conversions, advertisers do that...

Yes, it was a typo (I of course meant advertisers, as you can see in the other posts I wrote). Thanks for the correction.

and if it's accurate data like you claim, how come publishers can't see anonymous conversion data for every page? why can't we use it as a tool for improving our sites?

Good question. Why doesn't Google allow publisher to see the Smart Pricing status or conversion ratio? I don't know.

that's because you don't know what "smart pricing" really is...

Actually, we know quite a lot from the info Google has given us. Conversion is apparently one of the most important factors.
8:40 am on June 20, 2006 (gmt 0)

Preferred Member

10+ Year Member

joined:May 30, 2005
posts:456
votes: 0


don't annoy your visitors with ads they are obviously not interested in

Isn't it rather Google's job to try different ads to increase eCPM? I believe Google does that (rotates ads until it finds ads with the highest eCPM). Let's not forget that Google and publishers are in the same boat (they share the earnings).

By the way, a lot of people seem to emphasize CTR, while eCPM is obviously the most important figure. (Ads with 0.01% CTR and $50 eCPM still earn you more than ads with 50% CTR and $5 eCPM).

This 31 message thread spans 2 pages: 31