Forum Moderators: martinibuster
It's very very important to do a carefully controlled test of whether 3 actually helps versus 1. In many, if not most, instances, it actually hurts earnings because you "downsell", getting users to click on a lower EPC ad.
I see a number very important reasons why it is better to have no more than 1 block:
1. CTR will drop significantly because you double or even triple impressions while clicks increase insignificantly. There is theory that high CTR will affect smart pricing so that higher CTR will cause higher CPC.
2. Lower blocks will have less paying ads and part of clicks which otherwise would go to high paying advertisers will go to lower paying ones. This will hurt you CPC and CPM.
3. By limiting amount of adspace given to AdSense publishers overall encourage bidding war for top spots among advertisers. Just like OPEC keeps the West on short leash :)
However, multiply Blocks may be to the benefit of publisher in cases when you use 3 well placed one-ad blocks in right places on the page instead of one leaderboard, for example.
[edited by: fdmaster at 5:00 pm (utc) on Nov. 30, 2004]
Does it matter if your cpm drops? Does that have an impact on anything?
CPM shows how much money you're making for every 1,000 impressions. So yes, it's important (more so than EPC, for example), but the "your earnings" column is the number that matters most when you're paying for housing and groceries. :-)
but the "your earnings" column is the number that matters most when you're paying for housing and groceries. :-)
Some of us have other sources of income to pay for the latter. Also if you're working from home and use a room just for that (as most of us webmasters are) part of the housing cost can be written off against tax as a business expense. Course I think if you wrote off groceries you ate while working as a business expense they might start complaining about that. *grins*
In many, if not most, instances, it actually hurts earnings because you "downsell", getting users to click on a lower EPC ad.
I had that theory too. But after testing, the exact opposite was true.
In your case it may have been. I've tested it in a particular circumstance and downselling was exactly what happened. I had a carefully controlled test and waited for statistically significant results.
I would urge you to review carefully whether your test was controlled adequately and whether you had a statistically significant sample. My first day of testing led me to believe I might get a large increase by running three ads but when I had a statistically significant sample, the results were unambiguously negative.