Forum Moderators: martinibuster
Is there any truth to this?
I've always been more concerned with the bottom line. I have a few ads that perform very poor in terms of CTR, but they get tons of traffic monthly, so they typically make a good deal of money anyways.
I'm nervous about testing this theory, so I figured I'd just ask.
Would YOU remove a fairly well paying ad spot from your site simply because the CTR isn't great? Would this make enough difference to compensate for money lost by removing the ad (this would vary greatly by person... more of a rhetorical question)
Any input = greatly appreciated.
In the last 12 days, by removing some poorly performing ad placements, I have managed to increase the overall CTR of my primary site by about three percent over the course of 12 hours. It has held 3% higher since the ads were pulled from poor pages.
I will say that I have seen a small - but significant - bump in EPC. I can't attribute this 100% to the changes I made in CTR, but I made no other changes. I had no dips or hikes in traffic, ctr, or views.
I know that my version of significant may differ from yours - so we'll just say that significant for me, over a 10 day average amounts to a 30% increase (close to 4 cents)
My statistics are pretty steady... a 4 cent jump in EPC is abnormal for me... especially when I was in a period of decline for over 60 days.
Lots of things could have happened.... smart pricing... advertising changes... a million other things - but I'm inspired by this, and will continue to drop extremely low CTR pages in hopes that the trend upword continues.
Once again - I'll report back in a couple weeks.
The things that I have found work for me:-
1. Give each banner a channel, monitor it closely and pull it if it doesn't perform. I'm not sure if the banners that don't work will affect the stats or not, but when I ran two banners on one page, the main banner suffered badly as the second one got no clicks. Removing the banner restored earnings. Since that day I've only allowed an ad block to stay if it works. Visitors hate seeing ads on every page anyway, and become blind to them.
2. Experiment with ad style and placement. The Google heat map worked for me, but I found out about it the hard way by jiggling things around long before it was published!
3. Block ebay and all made for adsense sites. It's been my experience that advertisers selling goods and services to my niche convert well, and webmasters just trying to siphon off my traffic to click the same ads on their site don't. Blocking them has increased ny EPC six fold since I started blocking 2 months ago. Unfortunately the Google targetting bot makes a lot of stupid decisions on ad placement, despite having a lot of information on what ads convert and what ads don't at it's disposal. The only way to ensure you get quality ads on your site is to use the block list.
None of us really know how smartpricing works. That's no accident - Google don't want us to. However, I have one observation based on my experiences. Smartpricing hates change, and always reacts badly (downwards). Therefore, if you make sweeping changes such as blocking a lot of ads at once expect to see a downward turn in what Google thinks your site is worth. It took a couple of weeks, but it did recover and has been building ever since. If you are considering blocking advertisers, do it slowly!
You are right. There are far too many variables to attribute my mundane changes to one thing over another.
I am surely going to be sticking with the policy of removing poorly performing ads... site wide, or not - it seems to make some sort of a difference.
I agree with you 100% about the "change" theory. SmartPricing steps on my face every time I make a major change. I've been trying to make "baby changes" as to not disturb the smart pricing engine.
... hell, for all I know - my last 12 days trending could be the result of a slow recovery from a smart pricing reduction caused by changes made at that time.
Enough theory from me - back to what I know works... writing until my fingers fall off.
