A theory

[newbies]

Once in a while, you'd get contacted by a google account manager/optimizer/adviser, whatever, telling you they have some suggestions for you to boost your income. One thing they always offer is that you should put more ad units on your pages where you already have 1-2 units. You'd think that must be a good suggestion to follow. If you do, you would always see a drop in your overall EPC and income despite you may get more clicks by running additional ads on your pages. This happened to me at least 3 times and each time I ended up removing the additional ads suggested by the adviser.

From all these experiences, I come up with a theory: the so-called advisers are commissioned by google. They took from you a good chunk of your earnings from the change you made. That's why you earn even less after running more ads on your pages. They are greedy and thought all the clicks from your new channel belong to them. Correct me if I am wrong.

[incrediBILL]

They may make suggestions but that doesn't meant they'll work for you.

A/B testing of changes is the only way to make sure one works vs the other.

Enough with the Google paranoia already, just because you think they're after you doesn't mean you aren't being targeted ;)

[levo]

You get 68% from adsense for content..

[ember]

The advisor's suggestions worked for us. Bumped income by 20% to 25%. Maybe he was on commission, and if so, he won and we won. I am okay with that.

[netmeg]

Yea, in my case, the advisor did not suggest *more* ads (I was running 1 ad and 1 link unit on most pages, a couple pages had two ads) just different placement. And it worked. But when they sent me the mockup, they also had removed the link units, and I didn't take that advice. That site improved over 200% over all of 2010.

Google's never going to know your site or your users as well as you do (except algorithmically) I will listen to all the free advice they want to give me, but in the end, I will go with what makes sense to me. I don't mind testing new stuff (like for example when I allowed image ads back in) but it's still gotta at least potentially make sense to me.

[DanAbbamont]

How much your EPC decreases depends on how competitive your keywords are. The more competition there is, the less the difference between say the first listed ad and the 10th, even though that's kind of oversimplifying.

Also, how much time did you give it? It has to have enough data to figure out what ads to serve.