Plummeting CTR

is there a bottom

[Powdork]

Since April of 2006, my CTR has steadily dropped from the mid 4% range to the mid 1% range. Naturally this is a bit disconcerting, especially since the trend shows no signs of stopping. Has anyone else seen this, or am I the only one. Obviously there are many factors at work, but i would say that as a whole, I have pruned more low paying, low ctr pages than I have added new pages. For pages that have remained largely unchanged, including traffic sources, I am seeing similar trends. It's also important to note that the entire network does not see a whole lot of return traffic over extended periods. I should also note that i have not done a lot of testing to see what ad properties (colors, etc) work best, although i have tested with ad locations.
Anyone else (or maybe everyone else) seeing this?

[m0thman]

A year or so ago I was getting between 4% and 6% I get 4% on a good day and the average is around 3 to 3.5% I suspect it has more to do with crappy irrelevant ads than anything else.

[Hobbs]

4 to 1 % drop over two years is a huge drop, what's stopping you from testing colors and blending? There must be a decay in ad quality or content value or both for this to happen.

I have seen CTR fluctuations over time, but I'm more concerned with 1) Earnings then if there is something wrong I optimize for 2)eCPM of which CTR is just one factor.

[zett]

There is also the possibility that Google changes the way they count clicks (or better: the way they do not count clicks). The often abrupt changes I have been experiencing over the past 3+ years can not be easily explained by changes made to ad design or layout, by seasons or by bad targetting. I would not be surprised if there was something else.

Disclaimer: I know that Google can't do wrong, and that I am free to leave the program at any time for any reason if I am not satisfied with the program. I know that Google is said to "just protect the advertisers". I know. I know. I know.

[ken_b]

Has your traffic source mix changed much? If so, that could account for a good bit of the drop. Traffic numbers can be deceiving because the traffic sources can shift while the overall numbers remain fairly constant. And traffic sources can make a huge difference in CTR.

Overall, my CTR has been fairly consistent. But poor quality ads can send my CTR down drastically.

Mix a shift in traffic source and lousy ads, and things can get bleak in a hurry.

Do you use image ads? If so, how are they performing compared to text ads? Is one type doing much worse than the other. Has one type changed more than the other?

What about adlinks, do you use them? How are they doing compared to text and image ads?

[signor_john]

There is also the possibility that Google changes the way they count clicks (or better: the way they do not count clicks).

That might be one factor. In November of last year, Google announced that it was redefining the clickable area of an ad to save advertisers from paying for accidental clicks. Users now have to click on the ad's text link, not just on the ad as a whole. See Google's explanation [adsense.blogspot.com] of the change.

[purplecape]

Going back well BEFORE the change in the clickable area of the ads, I also experienced a gradual decline in CTR. Over the course of more than two years, CTR fell by half.

However, EPC has gradually risen and is now about double what it was, meaning that my earnings have recovered. There was a lag of several months, and you may not experience this, of course, but I offer this experience just to let you know that there may be a light at the end of the tunnel, and that CTR alone is not the only factor determining earnings.

[Powdork]

there have been no steep declines, just a steady drop of about .1% each month. it will probably rebound to about 2% in September/october as my summer traffic always brings a lower ctr.
i think part of the problem may be that most of my growth in impressions is a result of long tail searches. usually the user is finding specifically what they were looking within the content of the page. however, i am seeing the same trends on pages where the users are arriving via general searches and the ads direct them to the specifics.

[StoutFiles]

People do eventually gain "ad blindless"...if you shake the layout up a bit you'll see a dramatic boost in clicks.

I like to use a random number generator to mix up ad positioning in layouts.

//get rand number
if(num==1)
{layout1}
else if(num==2)
{layout2}
else
{layout3}

Mix it up; you'll be surprised what happens.

[zett]

That might be one factor. In November of last year, Google announced that it was redefining the clickable area of an ad to save advertisers from paying for accidental clicks.

No, no. That did not affect *me* too much. Visitors who click on my ads KNOW that these are ads (they are not blended and do not interfere with any navigational elements or typical user behaviour), and usually the targeting is good.

What I think we are seeing is that Google changes the way they COUNT clicks as they see fit, i.e. they might be saying - "ah, this click came from that user/network/IP address/country, so we think it is invalid and do not count it at all". It could also be a technical issue that their database could not capture all the clicks.

Fact is, that early in September 2007 all of a sudden Google began reporting LOWER clicks than my clicktracker (without having changed anything to the site, including the clicktracker and the reporting mechanism). This ended the two-and-a-half period of my tracker underreporting (in comparison to Google). Now my clicktracker was suddenly recording more clicks than Google did.

Interestingly, mid-May 2008 reversed things back again to its old state. Since then, the tracker is again underreporting.

The ONLY logical explanation I have for this is that Google did not count all the clicks between early September 2007 and mid-May 2008, or they did not count them as they used to do during the 2.5 year period before.

In numbers:

05/2005-09/2007 : +8.4% (Google reports on average 8.4% more clicks than clicktracker)
09/2007-05/2008 : -11.3% (Google reports on average 11.3% LESS clicks than clicktracker)
05/2008-now : +15.6% (Google reports on average 15.6% more clicks than clicktracker)

For those interested - the "Google grab" lasted from Saturday, 8th September 2007 to Saturday, 17th May 2008. I estimate that roughly 20% of the clicks were not counted during that period. Ouch.

* * *

(I think we had this discussion before and could not come to an agreement back then. Well, not exactly WE, but some members of this forum: [webmasterworld.com...] )