Would it be 21c less a penny, then increased by the ratio of the CTRs? ie might it cost 30c per click? This would seem to be the formula described in the FAQ but it seems unlikely that the ad in second place is paying more per click then my top ad. Or am I totally confused?
Does anyone know of any detailed documentation on how the cost is calculated?
UpDown
[adwords.google.com...]
In very brief detail, the bottom line is this: The AdWords system bids on your behalf to put you a tiny fraction above the Rank Number of your nearest competitor. And the smallest dollar amount it can use to do this is $0.01.
For ads on the right hand side, the factors considered are Max CPC and CTR, weighted equally.
For ads at the top, actual CPC is considered, and CTR has more weight than CPC. So you get to the top by being more relevant, more so than by paying more.
AWA
Particularly could you offer any more detail about the weightings and costs for ads at the top of the page? In my example in my first post is it possible to estimate how much the ad below me is paying if I assume what I believe are likely CTR figures?
Thanks
UpDown
...I am familiar with what is stated in the FAQ but I wondered if it was possible to be more precise about the formulae. Or are there any detailed examples?
Well, there was a detailed hypothetical example posted here quite some time ago. But I'll be darned if I can find it. It covered ads on the right side only, though. With all that said, the formula is very simple. It is simply your rank number as compared to that of your competitor.
Higher rank number prevails, of course. And again, rank number is equal to Maximum CPC multiplied by CTR.
So Max. CPC x CTR = Rank Number is the precise formula for ads on the right. Please note that this formula makes it possible for you to appear above someone who is paying more than you are.
Particularly could you offer any more detail about the weightings and costs for ads at the top of the page? In my example in my first post is it possible to estimate how much the ad below me is paying if I assume what I believe are likely CTR figures?
Sorry to say that that's about as much detail as I can provide regarding the Top Top spots. Not because I don't want to say more, mind you, but because the exact algo is not published - and I'm only aware of details in the general terms I've passed on here.
Sorry to disappoint!
AWA
Well, there was a detailed hypothetical example posted here quite some time ago. But I'll be darned if I can find it.
There are a few, although I still can't find the one I was looking for - this one is pretty good as well - detailed info by AWA:
[webmasterworld.com...]
Let me try it mathematically. Based on your example, your ad position is estmated to be 4.2 (0.21X20). The second position would have an ad position of a maximum of 4.19. Assume that Max CPC and CTR are weighted equally as the ads in the right side, the product of Max CPC and CTR should be 4.19.
Based on the limits:
Max CPC = 5 cents & max CPC = $50.0
Max CTR = 80% (100-20) & min = 1%
and CPCxCTR=4.19, the maximum and minimum values of CPC of the second position could be estimated.
For CTR = 1% the max CPC will be $4.19 (4.19/1)
For CTR = 80% the max CPC is 0.05 cents (4.19/80)
Thus, the max CPC for the second ad is estimated to be from 5 cents to 4.19 dollars.
UpDown
Seems like you were gone for a while there, no? I missed your excellent posts, and questions - and was glad to see your name appear again in the last week or so
;)
AWA
UpDown
I believe it is always based on the 'all time' CTR
UpDown is essentially correct, but here are a couple of additional details:
* The Max CPC considered is the Max CPC for the keyword searched in the moment the search is performed.
* The CTR considered is the CTR for the keyword in the moment the search is performed, taking into account the entire history. So this does in fact amount to 'all time' as UpDown suggested.
AWA
Seems like you were gone for a while there, no?
Sometimes you just need a vacation to get away from absolutely everything...
There is another post around here, that I can't find, that lists 2-3 bidding scenarios with CPC, CTR, multiple advertisers, and how bid rank is calculated based on how those advertisers change their bids, and how slight changed to CTR can affect a campaign - but I can't find it anywhere.
If someone comes across it - would be nice to see again to make sure the hypothesis in it are still correct with the more details we've learned from AWA over time.
That's another good example. The one I was looking for is actually one I wrote - which just makes it worse that I can't find it...
