channels, impressions and clicks

[bouche]

I finally setup channels after reading here that they are very important in deciding what parts of a site work and don't work.

I discovered, after making 3 channels that only one of the locations is really providing most of the clicks.

for instance there might be:

- 30 clicks on page A with 8000 impressions.
- 2 clicks on page B with 2000 impressions.
- 0 clicks on page C with 800 impressions.

Does the low clicks on pages B and C affect the overall earnings negatively? If I can't increase them, should I just remove them?

[MichaelCrawford]

I think it's best to remove ads from non-performing pages, but not to remove the pages themselves. The pages could still be drawing links and search engine referrals.

If you feel the mere existence of a page is harming your revenue, then if you choose to remove it, use a permanent redirect to send the traffic to a better page. Here's how you do it in apache. If you have just the one site on your server, then put it in the central configuration file. If you're on a web hosting service, put it in a .htaccess file in the same directory as the page you're removing:

Redirect perment /which-directory/removethis http://www.example.com/send-them-here.html

If you use IIS or some other webserver, say so and maybe someone else can help.

[lammert]

It is not always a good advice to remove ads completely from pages with very low CTR. There is no definite proof that they will cause smart princing to kick in. But when CPM ads are shown on your site, removing of the ads will certainly decrease your earnings.

I have one particular page which has a high impression count (almost 30% of the total site pageviews), but because of the nature of this page a very low CTR (1% of the AdSense clicks). In an email conversation I had with AdSense support about this page they specifically adviced me to keep ads on this page and allow CPM ads, to see how they performed as this might be a good example where CPM ads worked.

[MichaelCrawford]

What I was getting at is that ads might discourage links. If a page isn't the sort anyone would want to link to, then I would say leave the ads, but I think pages that draw links are valuable in their own right even if they don't produce revenue.

[bouche]

this is good advice. It makes me feel better about the google math.

I had 2 units (vertical and horizontal) on my highest visited page, and I noticed that when I took down the vertical linkage, the earnings went up. I assumed the obvious, that 2 views on one page with a click here and there didn't work out. So I naturally started thinking that pages with views, but no clicks would have the same effect.

Still, from alot of what I've read here, playing around and using the channels is the best way to try and get things moving up. I'll keep an eye on it for a week, and then maybe try it the other way for a week and see if there is a difference.

[MichaelCrawford]

What works for me, what works for that guy over there, won't necessarily work for you. And what works for you might not work for me either.

The reason is that our sites, and our visitors, are very likely very different. Some of my visitors don't like to click ads, but they do like to give me links, so I still work hard to write content that they will enjoy.

The best you can get from anyone's advice here are suggestions that you should test. Channels are an important component of that testing. But only testing your particular situation will yield the formula for your success.