It would be great if Google gathered together some of the stuff like the explainations you have given here in a Less Frequently Asked But Still Important Questions (LFABSIQ) document. There seems to be a number of questions that pop up that are beyond a simple FAQ and might cause confusion to the raw begginer if they had to wade through them, but the answers make the relationship much more profitable for advertisor and Google...
I think this is a great idea, and am starting this thread in the hopes of collecting feedback. What would be in this LFABSIQ document?
Let's see. I'm guessing we'd start with:
1.) Everything you've ever even half-way wondered about AdWords Algorithms....
(See, I do know you all...) ;)
AWA
I'm sure that many advertisers would opt for more targeted exact or phrase matched keywords if they understood more about them, but in the words of Tom Peters most of them go.....
Ready
Fire
Aim
Why are all the features (content network/search network/best ad rotation) opt-out rather than opt-in?
A lot of advertisers don't have a clue about where exactly their ads are displayed, e.g. on which content sites. As long as clicks are showing on reports many don't care. If they measured their ROI they may be in for a big shock!
I'm bringing this thread back up to the top at least this once, to request your feedback.
An "Advanced FAQ' is quite possible for the near future, and I'd very much like to have the comments of this august group, as to what should properly be in it.
Who better to tell me?
I'm compiling a list of suggestions for the product development folks, which is due on Friday. I'd certainly like to include yours.
Any takers?
AWA
I mentioned it before but I would like to know how the optimization mechanism works. I monitored it now for a while and see that within a few days it starts to favour a specific ad and serve almost all impressions there (which is what you would expect). However from time to time it seems keep trying the other ads.
I also would like to know more about the CTR; I'll be more specific I have a relatively high volume keyword which gets 1000's of clicks per month (phrase match, 2 words). Last week suddenly the impressions surged and CTR fell sharply. I added a few negative keywords and got things under control but what surprized me is that the keyword did not get stopped because it failed to have clicks on the last 1000 impressions. So you guys must have a "surge protection, we look at the Impression/CTR history", kind of algoritm.
I have another suggestion to. Why not organize a webcast sometime with an "advanced google adwords class". There are quite a few wizards around here and a lot of people that need to learn...
My 2cts.
I just saw something that had me stumped. How can Adwords report the following for a keyphrase:
2 Clicks
1 Impresstion
200% CTR
In the situation above, the account was charged for two clicks. These were the only reported stats for that keyphrase. It happened just as AdWords reported its first hits of the day to the account.
Does this question qualify for your LFABSIQ list?
Thanks!
* Explaination of how to run a champain using a general adgroup that contains a broadmatched term (Widget) and other adgroups for more targeted terms for different ads (Red Widget, Blue Widget).
* Explanation of the approval process, like when it's done, what causes an ad to be reviewed (random, changes, keyword, number impressions), the effect of needing approval, how to minimize downtime (timing of changes or creating new ads fo the changes and deleting old one later).
* The fact that CTR is total CTR, and many things seem to be based on the internal Google CTR.
* There seems to be a lot of threads on handling localization on Adwords and determining how to target and see them in other areas. I don't really follow those threads since I advertise US only, but just from the number, I would say that it is an area that probably needs more elaboration.
I'm a little mystified why sometimes my ad doesn't show up when there are only three or four ads on a page...Google seems to alternate them, I guess.
I'd like to know more about CTR...I guess it is straight forward, but how is CTR calculated across many keywords in a group, for example.
I have never run up against the infamous min bid set on some keywords for historical reasons...is there a way to know it is there.
I guess lastly, if you bid $x for a group with low and high value keywords, will the system only place you on the ones at or below your bid, or is it spread evenly.
Thanks...
I'm a little mystified why sometimes my ad doesn't show up when there are only three or four ads on a page...Google seems to alternate them, I guess.
I can already envision AWAs response to this one: it sounds like a daily budget issue.
If you're daily budget is set to $50, and it would take $100 to show your ads all day, then your ads will only show for 50% of the searches.
But it does kind of sound like a daily budget issue... ;)
BTW, thanks for all the posts to this thread. The more the merrier, as I compile a list to forward to product development on Friday.
I did want to make one quick comment though, about hudson's post:
I have never run up against the infamous min bid set on some keywords for historical reasons...is there a way to know it is there.
This is a thing of the past. At one point, each keyword had its own Min CPC. However, now, all keywords have the same Min CPC, which is $0.05 USD or equivalent in other currencies.
So much easier!
AWA
This is a thing of the past. At one point, each keyword had its own Min CPC. However, now, all keywords have the same Min CPC, which is $0.05 USD or equivalent in other currencies.
Ah...thanks for the clarification!
If you're daily budget is set to $50, and it would take $100 to show your ads all day, then your ads will only show for 50% of the searches.
ahh eWhisper...could be so, thanks
That is definitely not the case with me as I've set budgets for alll of my ads in the upwords of $10,000.
"Disappearing" ads and placement algo's should be much better explained. SImply put, if Google'd actually treat advertisers as customers/partners instead of as ppl that are out to "exploit" their algo's, then there'd be no doubt business would improve greatly.
I almost get a feeling of paranoia as it seems like google doesn't want anyone even close to knowing much about their processes before the IPO.
I think it'd be great to keep this thread alive. It's a good chance to have your feedback heard, for a feature that has a very good chance of coming to pass.
Thanks again!
AWA
