My 1 month experiment with the competitive filter

There's a kicker

[koan]

Getting tired of chasing after MFA sites and getting close to the maximum limit in the competitive filter once again (after emptying it a few times in the past and filling it back with the newer nuisances), I have decided to experiment with it and let Google do its thing, as they actually officially encourage us to do. I emptied my competitive filter. I've read some good things about finally letting it go from some users on this site. Still, on bad days it was difficult not to cancel the experiment but I needed to figure this out and I knew I had to give Google a few days to try out some advertisers and calculate their quality scores.

So on January 18 2008, I emptied my competitive filter. But here's the kicker that actually makes this experiment somewhat more scientific. I share 3 sites with a partner where I do the webmastering work and my partner produce content. Adsense is displayed 50/50 using a php random function. I didn't empty his competitive filter, which was a copy of mine, as I take care of this kind of stuff with his adsense account, being the web guy and him, the writer. Let's look at the result from January 18 to February 15 2008 (yesterday).

Site A:

Description: high traffic (for me), low competition among advertisers. Hence general low CTR, low EPC, but making it up with volume. Somewhat. Hasn't done too well since october.

Impressions represent 49.83% of total of impressions for the 2 accounts.
Clicks represent 51.55% of total of clicks for the 2 accounts.
Earnings represent 49.28% of total earnings for the 2 accounts.

My conclusion: The CTR was higher, but as goes the old saying on this forum, the net earning was lower. The difference in earning for me is acceptable but could be of concern with a very high traffic site.

Site B:

Description: average traffic, high competition among advertisers. High CTR, high EPC (relatively to my sites in general, some of you would probably scuff with your super duper 10 years old authority sites on health services or stock market tips, but anyway).

Impressions represent 50.12% of total of impressions for the 2 accounts.
Clicks represent 49.07% of total of clicks for the 2 accounts.
Earnings represent 49.75% of total earnings for the 2 accounts.

My conclusions: Lower clicks and slightly lower earnings. The difference in earning for me is minimal.

Site C:

Description: low traffic, average competition among advertisers. Average CTR, average EPC. Quite the mister average in most ways. Statistically speaking, this one may be of lower interest because the small traffic makes the data more chaotic.

Impressions represent 50.36% of total of impressions for the 2 accounts.
Clicks represent 41.19% of total of clicks for the 2 accounts.
Earnings represent 48.55% of total earnings for the 2 accounts.

My conclusions: Way lower clicks and slightly lower earnings. The difference in earning for me on this one is slim to none.

Final thoughts:

Of course, this doesn't account for possible discrepencies in smart pricing level for both accounts. So far, the difference in total earnings is somewhat minimal, but could possibly of concern for very high traffic sites where 1-2% means hundreds of dollars. However, for someone in my position, I did avoid wasting time chasing useless advertisers instead of working, which makes the difference in my final decision to stay the course with an emtpy competitive filter. My only worry, however, is bad user experience (especially if some advertisers are installing malware) and a diminishing trust over a long period of time with the ads, and the Google brand. But if I calculate the time saved versus the earning lost, I do end up a winner in the end.

As for my other sites, it's difficult to tell how they were affected since too many other variables come into play (seasonal, traffic change, etc)

Thanks for reading.

[skweb]

Thanks for sharing the results; mine list is empty too because I too learned that if you add one domain to the list, a 100 more are there waiting to show up on your website. It is much better to focus on building your website so that Google algorithm raises the price for your website and driving away the crooks.

[vivalasvegas]

Interesting, thanks.

[europeforvisitors]

I use the filter for its intended purpose (as a "competitive ad filter"), and I've got only a handful of domains listed. I just don't have the patience to pick and choose ads or advertisers, especially with geotargeting making the whack-a-mole task more complicated than it would be if the same ads were served worldwide.

[koan]

EFV, with the adsense preview tool, I limited myself to checking US and Auto geotargetting lists, as that would represent 80% of the total traffic on my sites. It's my opinion that most undesirable MFA sites don't target geographically and go in auto mode, so I think I got most of them there (I could be wrong).

I think MFA advertisers affect mostly sites with low competition among advertisers where they can actually compete with good CTR ad minimal EPC.

[Jane_Doe]

I just block ads that are off topic or are exceptionally spammy. Like on one site I used wording something like an elegant "affair" to refer to a celebration and I ended up with divorce lawyer ads. It took me a while to figure out what was triggering all the divorce stuff, then I realized it was the multiple meanings of the word "affair".

Or sites with URLs like www.buy-really-cheap-meds-here.com or www.make-money-and-lose-weight-while-you-sleep.com. I don't want ads like that on my site no matter how much they pay per click.

[edited by: Jane_Doe at 7:33 pm (utc) on Feb. 17, 2008]

[greatstart]

I had to dump my filter too this month. Despite all the MFA sites coming back, my CTR has gone up again. I just hope my visitors don't get too upset seeing the same run-of-the-mill ads all the time.

[inactivist]

My particular nuisance is ringtone ads. I gave up fighting, too, and cleared my filter of all but two of the most offensive advertisers, none of which were ringtone advertisers.

Result: Revenues are up, CTR is up, eCPM up, EPC up. Go figure.

Lesson learned: Stop fighting market forces and work *with* them.

[farmboy]

So how's the experiment coming along now koan?

FarmBoy

[koan]

Hi farmboy, I have emptied my partner's competitive filter and although it did seem to have reduced my earnings a little bit, the time gained in not micromanaging this stuff is worthwhile enough, so I'm staying the course...

However, I'm thinking of doing experiment #2, which would be adding back my old list on my account and leaving my partner's account empty and see if there's any difference again. That could confirm my conclusions.

[potentialgeek]

I have very little respect for the Adsense leader at Google, Inc. who refuses--and this has gone on for years--to provide not just a fraud filter, but no justification for not providing it.

To my knowledge, Google Inc. hasn't even released one case study, one success story, etc., showing how one publisher flushed out a Competitive Ad Filter and their income went way up, and their brand wasn't defiled by cheesy, dishonest, and irrelevant ads.

To send out automated monthly memos referring only to the potential for higher income, with no respect for the significance of brand integrity, is pure condescension.

So thank you for sharing your experience. Will Google be strong enough to follow your lead?

You know the weakness of Google's effort to convince us to empty the filter is probably because it simply doesn't have the data to show it would make much difference if any to the bottom line.

If they thought our income would double, do you think we might get more than an automated email?

p/g

[annej]

I just block ads that are off topic or are exceptionally spammy.

Me too. I don't filter out ads to make more money, I filter them out to make more of my visitors happy with my site.

[farmboy]

However, I'm thinking of doing experiment #2, which would be adding back my old list on my account and leaving my partner's account empty and see if there's any difference again. That could confirm my conclusions.

I emptied my filter and haven't observed any negative results.

FarmBoy

[htaccess]

> koan

by reading your interesting topic another question comes up.

Is it possible and OK within the TOS to run 2 adsense accounts on one site? (Not on the same page, that's clear). I have since years a company adsense account and my wife has a private adsense account. We are running adsense on total different web projects. For tax reasons I would like to have on some months more income on the private account than on the other. Is this OK and approved?

Many thanks

htaccess

[koan]

htaccess, I won't go into details as it is off topic, but if you can program a script that will make sure only one account's ads are displayed on a page at a time, there is no problem with Google. Revenue sharing web sites are based on that principle. The random and include functions with PHP would provide all you need. I think it's a great way to work as a team on a site if you know someone who has a lot to offer in terms of content and you're web savvy (or vice versa), without too much messy tax stuff in the end.

However, if a site gets one adsense account banned, the other one will too.

[RonS]

I may have read it wrong, but didn't the most recent AdSense terms remove the restriction on multiple accounts on one page?