32nd Post
Posted 4th March 2009 at 03:46 PM by Zale
dkruglyak wrote: Zale, Looks like you got most of the issues addressed and quality scores are the proof... I thought that might be the case as I posted only after reading Page 1 of this thread.
I apologize for the delay in responding. I had written a wonderful response yesterday. But didn't push submit. Then when I discovered it today - I had timeed out.
I am lucky I am good looking - cause I am not so bright.
Anyways thank you for the detailed response. Let me respond to (AGAIN) all your comments
dkruglyak wrote:
I have just a few minor comments:
Not so minor - very helpful!! Please keep them up. FYI - some real experiences with numbers would help. it makes it less theoretical. ie; QS of 10 was perfect.
dkruglyak wrote:
1) Does EVERY ad in EVERY ad group matches keywords? Any bad combinations would drag down quality scores of keywords, ads and the account.
Yes. Originally the keyword was the headline. I have now gone through many of the low QS scored keywords and adjusted the ads. I have tried to ensure that at least one of the keywords is in the headline.
dkruglyak wrote:
2) My opinion is that running 10 ads per adgroup is an overkill. Would take you 10 times longer and costly to collect statistically significant data. I would run just 2 or maybe 3 ads at a time and focus on best CTRs, before focusing on conversion of clicks that come from these ads.
I agree - I am just not sure what to drop right now.
There is an equal distribution among all the ads.
I realize if I reduce the number of ads - I will get a better data about which keywords are drawing.
On the other hand if I cut down the ads. I may remove an ad that attracts buyers and promote tire kickers. And with the little sales (1 !!) I wont' know.
I spent quite a bit of time developing the ad content.
And I used ad content that was provided by my advertiser who said his ads were based on his experiences.
Any ideas or thoughts on how to select the ads other than through what drives sales would be useful.
FYI - Since I removed the search partners - I have not spent my budget. This is consistent with my experiences to date. It takes google 3 or 4 days to hit budget spend after a major change.
dkruglyak wrote:
3) You still might have to atone for your account history! Just focus on ad CTRs to get on Google's good side and get your quality scores above 7 and 8, hopefully to 10 so you could reduce click costs.
I am working on it.
Some of the things I am doing with the low score items:
(Using all time average for this)
1) Increasing CPC max for ads not displayed by .20
2) Increasing CPC max for ads in position > 8 by .10
3) Adjusting Headlines by changing the headline from the EXACT keyword - to a more human readable phrase incorporating the headline. For ads with QS < 7 - Once I have adjusted the headline I leave them alone.
I don't want to adjust the ad - cause it will screw up my T202 tracking too much. If my ad is SO different than t202 - then t202 reporting will be useless. I am assuming that the 2 lines are as important as the headline when a person decides to CLICK.
4) I can't think of anything else. Suggestions welcomed!
dkruglyak wrote:
4) Your post makes it appear as if you are "mechanically" copying the keyword into the ad and LP headline. Just to be sure the flow needs to make sense to humans, not just to satisfy Google's QS algorithm.
Yes that is exactly how I started.
I am now doing that next step. Its easier to use the QS to help decide which ones need to be changed.
Thats due to your suggestion by the way.
dkruglyak wrote:
5) Search partner traffic could be valuable too. I just suggest running it in a separate campaign, using different bids and gathering / analyzing different stats. Like which search partner brings what clicks and what conversions.
Yes. I now have removed Search partners and Content search totally out of the system. I realizing its adding to many variables to the mix.
I think each program has to be run separately and differently.
dkruglyak wrote:
6) Sure, people like to research (multiple clicks) before buying high-ticket items. Help them with you LP sales copy - offer "free reports" in exchange for email address and keep sending them emails till they buy something!
To simplify my getting to profitability - I am using my advertisers landing page. Putting my own LP into the mix - will add one more variable too much. I will be adding my own LP - but just later.
dkruglyak wrote:
Happy to bounce around these ideas, I learn from this too... What are your conversion rates and EPCs so far?
[/quote]
I have only made a single sale from about 192 Adwords reported Clicks. T202 reported 267 Clicks in the same time period. T202 tends to over report - but moves closer to the google number as time marches forward. FYI - Clickbank reported 3,233 Unique in the same time period. A number close to the 11,503 ALL clicks reported by T202.
The program has been running for just about a week. And I removed the Search partners three days ago.
My daily budget is only $25.00
I apologize for the delay in responding. I had written a wonderful response yesterday. But didn't push submit. Then when I discovered it today - I had timeed out.
I am lucky I am good looking - cause I am not so bright.
Anyways thank you for the detailed response. Let me respond to (AGAIN) all your comments
dkruglyak wrote:
I have just a few minor comments:
Not so minor - very helpful!! Please keep them up. FYI - some real experiences with numbers would help. it makes it less theoretical. ie; QS of 10 was perfect.
dkruglyak wrote:
1) Does EVERY ad in EVERY ad group matches keywords? Any bad combinations would drag down quality scores of keywords, ads and the account.
Yes. Originally the keyword was the headline. I have now gone through many of the low QS scored keywords and adjusted the ads. I have tried to ensure that at least one of the keywords is in the headline.
dkruglyak wrote:
2) My opinion is that running 10 ads per adgroup is an overkill. Would take you 10 times longer and costly to collect statistically significant data. I would run just 2 or maybe 3 ads at a time and focus on best CTRs, before focusing on conversion of clicks that come from these ads.
I agree - I am just not sure what to drop right now.
There is an equal distribution among all the ads.
I realize if I reduce the number of ads - I will get a better data about which keywords are drawing.
On the other hand if I cut down the ads. I may remove an ad that attracts buyers and promote tire kickers. And with the little sales (1 !!) I wont' know.
I spent quite a bit of time developing the ad content.
And I used ad content that was provided by my advertiser who said his ads were based on his experiences.
Any ideas or thoughts on how to select the ads other than through what drives sales would be useful.
FYI - Since I removed the search partners - I have not spent my budget. This is consistent with my experiences to date. It takes google 3 or 4 days to hit budget spend after a major change.
dkruglyak wrote:
3) You still might have to atone for your account history! Just focus on ad CTRs to get on Google's good side and get your quality scores above 7 and 8, hopefully to 10 so you could reduce click costs.
I am working on it.
Some of the things I am doing with the low score items:
(Using all time average for this)
1) Increasing CPC max for ads not displayed by .20
2) Increasing CPC max for ads in position > 8 by .10
3) Adjusting Headlines by changing the headline from the EXACT keyword - to a more human readable phrase incorporating the headline. For ads with QS < 7 - Once I have adjusted the headline I leave them alone.
I don't want to adjust the ad - cause it will screw up my T202 tracking too much. If my ad is SO different than t202 - then t202 reporting will be useless. I am assuming that the 2 lines are as important as the headline when a person decides to CLICK.
4) I can't think of anything else. Suggestions welcomed!
dkruglyak wrote:
4) Your post makes it appear as if you are "mechanically" copying the keyword into the ad and LP headline. Just to be sure the flow needs to make sense to humans, not just to satisfy Google's QS algorithm.
Yes that is exactly how I started.
I am now doing that next step. Its easier to use the QS to help decide which ones need to be changed.
Thats due to your suggestion by the way.
dkruglyak wrote:
5) Search partner traffic could be valuable too. I just suggest running it in a separate campaign, using different bids and gathering / analyzing different stats. Like which search partner brings what clicks and what conversions.
Yes. I now have removed Search partners and Content search totally out of the system. I realizing its adding to many variables to the mix.
I think each program has to be run separately and differently.
dkruglyak wrote:
6) Sure, people like to research (multiple clicks) before buying high-ticket items. Help them with you LP sales copy - offer "free reports" in exchange for email address and keep sending them emails till they buy something!
To simplify my getting to profitability - I am using my advertisers landing page. Putting my own LP into the mix - will add one more variable too much. I will be adding my own LP - but just later.
dkruglyak wrote:
Happy to bounce around these ideas, I learn from this too... What are your conversion rates and EPCs so far?
[/quote]
I have only made a single sale from about 192 Adwords reported Clicks. T202 reported 267 Clicks in the same time period. T202 tends to over report - but moves closer to the google number as time marches forward. FYI - Clickbank reported 3,233 Unique in the same time period. A number close to the 11,503 ALL clicks reported by T202.
The program has been running for just about a week. And I removed the Search partners three days ago.
My daily budget is only $25.00
Total Comments 0
