7/10 QS more expensive first page bid estimate than 5/10?

[dave1814]

Hi Guys, hoping someone can help me...

Basically im testing different ads with the same keyword and found out (after quite a bit of time!) that the ad that rates the keyword as a 7/10 QS is £19 first page bid estimate and the ad that rates the keyword a 5/10 QS has a £11 first page bid estimate!

Why is this and which should I keep?!

thanks,
Dave

[smallcompany]

Can you explain how you test? Two ad groups with same keywords, or same ad group, or...

[Baylow]

I just saw this happen as well, but cannot reproduce it. I have a theory on why it might happen (this is TOTAL conjecture and i have no evidence to back it up):

Every time a user does a query that matches to your search term Google calculates what it would take to be on page one for that query. If you average all the possible queries you've potentially shown for (and then do some rounding) you come up with that keyword's EFPB.

Now, when your quality score jumps from 5 to 7 you become eligible for more user queries. These user queries are higher quality, and most are on Google.com vs. the Search network. Due to the higher quality of the traffic the bids and competition level is higher. This causes FPB to increase.


*edit typos.

[outland88]

I'm beginning to think both Adwords and Adsense are more broken than Google will ever admit to. I’ll guarantee you Google’s best engineers couldn’t tell a federal investigator what the price of anything would cost in the Adwords system with any degree of consistency. I think it will become even more topsy- turvy with Google screwing up the Search results so much people are being forced into Adwords.

[smallcompany]

I’ll guarantee you Google’s best engineers couldn’t tell a federal investigator what the price of anything would cost in the Adwords system with any degree of consistency

No head, no tail.

Look what's happening in Italy. [webmasterworld.com...]

Google works better than Fed's green printing machine.

[Channel01]

I honestly wouldn't place much stock in the first place bid estimates. I've seen way too many odd occurrences with the estimates to use them for anything more than a rough guide, at best.