Forum Moderators: martinibuster
My site layout does not give me much flexibility in placement, so my first option has been to place the ads just below my content. Unfortunately this is also just below the fold of the page and theirfore I'm affraid that visitors may miss seeing them at all on some pages.
Has anyone else used a similar placement and had good results? Should I shoot for a placement above my content even if it may not be as 'pretty'?
Granted that we added the codes to more pages as we went along, we found our revenues jumped 5x the following month when we decided to junk the below-the-fold position and focused on putting the code either as a leaderboard at the top of the fold or as a skyscraper.
There's a reason why banner ad networks only pay 0.05 cents CPM rates for below-the-fold banners as against $1.00 CPM (in some cases lower or higher) for top-of-the-fold banners: greater visibility means better performance.
In order to quantify what affect can be attributed to change in ad placement, CTR is what should be compared, not revenue. Can you estimate the effect on CTR on your site? If for example your CTR before the change was 2% and it increased to 3% it shouldn't be a TOS violation to share that your CTR increased 50%.
Placement below the fold makes more sense when a large percentage of visitors will read enough of the content to scroll to the bottom. If a site has long articles it's often worthwhile to consider splitting it across multiple pages to minimize what's below the fold and ensure ads are always (or mostly) visible on page load and after scrolling.
I understand that CTR is influenced by a myriad of factors. In our case, the only change we made in July to August was to change the placement of the code - from end of article to top of the fold. Hence, the significant jump in revenues can be attributed solely to the change in placement.
I've tried both leaderboards and skyscrapers on my travel-planning site, and I haven't witnessed any statistically significant differences in CTRs or revenues when switching back and forth. But then, I have an editorial "content site" with most of its traffic on inside pages that readers have found through search or by drilling down through my table-of-contents pages. So any benefit to having a leaderboard on top (where it would be seen immediately) may be offset by the benefit of having a skyscraper next to the editorial content that users have made an effort to find and read.
I may experiment with the leaderboard format again, but in the absence of compelling evidence that it will perform better on my site, I prefer to use the skyscraper and avoid the "too many ads" look that one finds on corporate sites like Travelandleisure.com, Concierge.com, and the travel pages of general news and entertainment sites like MSNBC.com. Having all of my ads and affiliate links in one place (the right column) makes for a neater layout and doesn't send a subtextual message of "We're trying to make money and editorial content is an afterthought."
The way I see it, editorial quality is my stock in trade, and avoiding a commercial "weekly shopper" is one way to inspire reader trust. That in turn should lead to more affiliate sales and more AdSense clicks by readers who are serious prospects and not just information-seekers. (And no, I'm not trying to be charitable toward advertisers--I just want them to be happy with the results of "content ads" so they'll continue choosing that option in AdWords.)
Maybe it will just take some tinkering to get things to work with my layout and visitor combination.
CTR is important - but I'd rather focus on revenues.
Don't get me wrong - revenue (or CPM) is what it comes down to at the end of the day. But revenue is affected by so many factors that if you truely want to know whether a change in ad placement, color scheme, etc. is effective you must focus on CTR. And by CTR I mean *real* CTR, not the CTR reported by the AdSense reports. Real CTR is clicks divided by potentially paying impressions, meaning impressions consisting of PSAs and alternate ads must be subtracted out. Since it's *impossible* to measure PSAs that means alternate ads must be implemented to make accurate comparisons.
In our case, the only change we made in July to August was to change the placement of the code - from end of article to top of the fold. Hence, the significant jump in revenues can be attributed solely to the change in placement.
At a minimum, adding the code to more pages contributed to the increased revenue. A change in EPC could have too (or could have had a negative effect). You mentioned moving to two different ad formats so if your mix of ad formats wasn't the same before that could have had an impact.
I'm not disputing that changing the ad placement had a positive impact. But based on what you shared it may not have been responsible for the full 400% (5x) revenue increase you described.
Let's take a hypothetical site which had its revenue increase 5x between Period 1 to Period 2. Assume 50% impression increase is partly due to adding AdSense code to more pages and partly due to more traffic to the site.
Period 1:
Ad format - leaderboard below fold
Clicks: 10
CTR, Reported: 1.00%
CTR, Real: 1.43%
EPC: $0.20
Daily Impressions: 1,000
Daily Real Impressions (Excludes PSAs and Alternate Ads): 700 (30% PSAs/alternate)
Daily Revenue: $2.00
Period 2:
Ad format: leaderboard above fold
Clicks: 32
CTR, Reported: 2.13%
CTR, Real: 2.67%
EPC: $0.25
Daily Impressions: 1,500
Daily Real Impressions (Excludes PSAs and Alternate Ads): 1,200 (20% PSAs/alternate)
Daily Revenue: $10.00
Revenue increased 400%, but it was partly due to increased overall impressions, partly due to a higher percentage of impressions showing paying ads, partly due to increased earnings per click and partly due to increased CTR. Since the ad type (leaderboard) was consistent we can rule that out as a factor in the CTR change and to keep the example simple we'll assume no color/formatting changes or improved ad copy of the ads displayed. That means CTR increeased 87% ( ( 2.67 - 1.43 ) / 1.43 * 100 ) due to moving the ads from below the fold to above.
Revenue increased 400%. CTR increased 87%. Moving the ads above the fold was a good decision, but it wasn't responsible for a 400% jump in revenue since the bulk of that jump was due to other factors. That distinction might not seem important, but if you want to analyze the real effect of changes and make smart informed decisions it is important to measure the right metrics. Otherwise, not only could you be under/overestimating the impact of a change, you could actually think a change that is really positive is negative and vice versa.
Dave.
Dave.
Have you tried moving the skyscraper to the top of the page or moving a small amount of the content above it so that it's at least partially visible above the fold? You could always try this type of layout for that column:
Current top of column content minus 2 listings
Note that more listings below
AdSense skyscraper
More listings
Richmondsteve....You're pretty gutsy posting your Adsense stats
I appreciate the warning and agree that violating the TOS is a no-no. You must have missed where I wrote "Let's take a hypothetical site..." I just made up some stats to show that a 5x revenue increase (which is what alika mentioned) can be impacted by a lot of variables and changing one of the variables while ignoring the impact of the others can be misleading.
As far as placement is concerned, I suspect (but I can't be sure, 'cos Google doesn't offer the detailed stats) that by making the ads appear across the fold (on one site) made a reasonable difference to my earnings. That is to say that I have a site that has many pages that are about 2 screen depths long. I placed the ads on the right hand side, but exactly half way down, so no matter where the user is some of the ads are visible. Any pages that I had that were too long for the ads to be visible at any point, I modified them accordingly.
I think the optimal placement is totally dependant on the type of site you have. I have AdSense running on a few sites now and each one has it's own characteristics and AdSense compliments each one in different ways. The formula that works for one site doesn't necessarily work for the others.
2odd...
Have you tried moving the skyscraper to the top of the page or moving a small amount of the content above it so that it's at least partially visible above the fold? You could always try this type of layout for that column
I tried putting the Google AdSense skyscraper between my top-performing affiliate links and some of the lower performers, but that didn't work very well. I've forgotten exactly what the problem was, but I think I was seeing a big gap between the bottom of the AdSense box and the affiliate links.
I did get rid of some marginal affiliate programs when I added AdSense, and I relegated others to links on specific pages, just to keep AdSense from being too far down on the page.
richmondsteve -- Yes I know what affects CTR. But for the purposes of the original query on this post and in reference to my response, we did not do any of the above to see the jump in revenues -- only the change in ad placement from below the fold to above the fold. No color scheme change, no change in ad layout.
So as I earlier stated, we found that the above the fold works much better than below the fold. And this is from our experience where we put the codes at the end of the articles (often from 500-2000 words long).
On one of my sites my revenue fell about 50% from one month to the next which coincided with a move from banner to skyscraper. But it turned out that the revenue drop was due to a 75% drop in EPC and an increase in percentage of PSAs (and alternate ads) from 10% to 80% offset by adding AdSense to more pages (more impressions) and increased *real* CTR. Real CTR (again, not what is incorrectly labeled as CTR in the AdSense reports) more than doubled. If I didn't factor that in I could have mistakenly determined that the move to skyscraper was detrimental, while in fact it was very positive.
Anyway, hopefully some readers that didn't understand the effects of variability and need to isolate a given variable when evaluating changes do now. There's nothing more for me to add.
#1. Change in daily paying impressions between period 1 and period 2. You yourself noted "Granted that we added the codes to more pages as we went along".
#2. Change in EPC between period 1 and period 2.
Those are probably the biggest factors that need to be taken into account. Whether such factors need to be taken into account to understand the effect of your change is not really debatable, it's fact - unless you know that EPC remained constant and revenue per paying AdSense impression remained constant (odds of that are slim to none).
#1 can be derived from the AdSense reports after subtracting impressions that are PSAs (which you can't track) and alternate ads (which can be tracked if you know how). #2 can be derived directly from the AdSense reports.
My explanations for why these factors need to be removed from the equation to have any chance of accurately measuring the effect of the change in placement and ad layout (you did say period 1 was some undisclosed type below fold, period 2 was combination of leaderboard above fold and skyscraper).
Anyway, I already said I had nothing more to add and I've now made essentially the same points in 3 posts to this thread. I sense this is being taken as confrontational and critical when I was simply trying to point something out that you appeared to be ignoring and that others might ignore as well. I don't want to keep repeating myself so unless I can contribute something of real value to this thread this will be my last post to it. Feel free to have the last word.
Don't know if this is true or not but I think the eye tends to drift to the left
