|Scared to experiment and create new campaigns.|
Through fear of disrupting my profitable ones!
| 4:14 am on Sep 11, 2007 (gmt 0)|
Just wondering if anyone else has come across these feelings.
For almost 2 years now I have been running a couple of very successful AdWords campaigns. I've recently started a couple of new projects, and wanted to buy AdWords to promote them, but I am finding myself hesitant through fear of reducing some kind of "Account Wide" quality score and disrupting my good, profitable campaigns.
When I do pluck up the courage, I go through a predicable cycle of rapidly deleting any keyword that doesn't achieve 1% (I don't even give myself a chance to experiment with the ad copy) before inevitably deleting the new campaign all together.
I understand the basic idea and need behind Quality Score, but I can't help thinking that Google have shot themselves in the foot a little because I can't be the only one who's spend with Google is now throttled by some kind of irrational fear that they have setup in my mind!
Is one strategy to create a second, experimental account, and then move successful campaigns into my main account once i've got them working?
| 3:12 pm on Sep 11, 2007 (gmt 0)|
I'm not aware that AdWords does any account-wide quality scoring. I run different campaigns to different sites, all within one account. Some sites have great quality and I can run really profitable ads to those sites. Other sites have been hit with a lot of high minimum bids due to poor quality. I can still run some keywords to those sites, but they're not nearly profitable as the other sites. But I don't see any effect those low quality keywords have on the campaigns with the high quality keywords.
| 4:05 pm on Sep 11, 2007 (gmt 0)|
I think maybe i'm worrying too much (common trait of mine!)
In a Campaign A, i'm bidding on:
...both of which are relatively low volume, however "widgets" on it's own is a high volume term. I created a new site, related to the same "widgets", and wanted to go for the generic traffic, so in "Campaign B" I bid on
(the same widgets as in Campaign A) and at the first statistics update it had over 1000 impressions and zero clicks. I broke out in a cold-sweat and deleted the campaign, having convinced myself that the quality algo would somehow hit Campaign A as a result.
My spend on Campaign A is about $2000/mo, so I would hope that Google wouldn't do anything too rash - but I know that even that spend is insignificant as far as Google is concerned, but represents my livelihood as far as I am concerned.
I really want to spend more with Google, but I daren't!
| 4:40 pm on Sep 11, 2007 (gmt 0)|
This is an important topic, dmorison, and I thank you for starting the thread.
I'll pass the link of this thread along to the right folks later in the week (late Thursday to be exact, as a 'top thread' in the WebmasterWorld section of the Advertiser Feedback Report) - and I'd be very interested to hear how others feel about this, regardless of which side of the fence you might fall on.
| 4:50 pm on Sep 11, 2007 (gmt 0)|
Many thanks, AWA!
I've subscribed to adwords.blogspot.com and look forward to any nerve carming updates!
| 5:21 pm on Sep 11, 2007 (gmt 0)|
I have some *very* diverse campaigns in my account (and the accounts I maintain for clients) - some of which were raging successes, and some of which were raging failures. I really haven't seen where a bad experiment can adversely affect existing campaigns that are running well - and I've been doing this practically since there was an AdWords.
I suppose it's possible (because I only have my own examples, and who knows, maybe they WOULD be even better if I hadn't experimented in the same account) but out of all the things there are to worry about, I don't generally worry about this one.
| 5:27 pm on Sep 11, 2007 (gmt 0)|
|This is an important topic, dmorison, and I thank you for starting the thread. |
AWA, please do let them know.
I'm in the same position.
I have a bunch of accounts in MCC, with each account targeting one product/site/project. That lets me keep financial accounting separate.
Here is an example.
I have an account where I only spend $1000/mo on ads and make $3500/mo in profit. The account is just to promote one product. I've been running it profitably for years.
Now, I know I could substantially increase spending by adding more keywords and restructuring content campaigns for tighter theme targeting. I would drop keywords that don't drive any traffic and won't put useless strain on your system. I would be making more sales, Google would be making more from my ad spending. Life would be great.
Yet, I'm not even touching that account because it's been running that way for a couple of years and I'm afraid to touch it.
I changed my tracking system six months ago (I use an in-house one, not Google Analytics), but I still have old urls in that account pointing to simple redirects on my domain that no longer do any tracking, just redirect to the landing page.
I'm afraid of even changing the creative and keyword-level urls to new tracking.
I've had accounts that lost all traffic overnight back when the search and then content quality scoring was implemented. So I'm afraid to change ANYTHING with the accounts that weren't affected.
And that's the situation I have with many of my older accounts that are profitable. They could be made more profitable, but it's better to be making something that making nothing at all. So I don't change anything.
I think this might be one side-effect Google didn't anticipate.
I really feel pathetic sometimes, but that's the reality of my life as Google advertiser.
| 6:06 pm on Sep 11, 2007 (gmt 0)|
I suffer from this. We have only removed stuff for months. The big fear is to get a quality hit to a major domain and lose existing campaigns. Also any new campaign is not as likely to settle in a manner like they did in the past so it is harder to absorb the initial high costs.
| 1:22 am on Sep 12, 2007 (gmt 0)|
I have many adgroups that could use updating. I also would like to change some of my urls but am afraid. My spend is high, but would definitely be higher if I weren't worried about tripping some kind of editorial process that would destroy my campaigns. It is strange to be so afraid even when you know you aren't doing anything wrong.
It is all a product of Google implimenting campaign-killing changes (ie. superhigh min bids), without telling advertisers what they were doing wrong. Just knowing that I could be 'slapped' and never told why or how to fix it, I worry that any change I make could be the thing that brings me to ruin.
More information about quality score and how to improve or fix landing pages (specifics) would go a long way towards easing the minds of advertisers. We are grasping in the dark, and that is, well, scary.
| 2:05 am on Sep 12, 2007 (gmt 0)|
Fear is usually caused a lack of control or a feeling of lack of control. I certainly have that.
| 3:41 am on Sep 12, 2007 (gmt 0)|
Don't be afraid to experiment at all, that's the only way to find the "sweet spot". However if there is one point of advice I could give every person who struggles with out of control spending in AdWords it would be, ditch broad match. Just don't use it at all. Turn it off, use only phrase and exact. In fact, you might just start with exact if you have a very limited budget. If using phrase then keep your eyes constantly on the search query report and build up a list of negative keywords to keep things on topic. Only after you have an extensive negative keyword list should you venture into the world of broad match. And then check your query reports as often as possible. Manage your negative keywords just like your positive ones, always researching, always building.
AdWords' over assuming broad matching technology is probably the single most damaging factor to a new campaign's cost effectiveness and should be used only after you master the mechanics of exact match and phrase. It would be nice if Google started you out on the path to success instead of pointing your cart directly into a downward spiral from the beginning, but I guess they'd rather get a quick fast burn of cash from countless dupes instead of building long term successful relationships. Probably not the case, but their initial account settings sure seem to indicate otherwise.
| 3:32 pm on Sep 12, 2007 (gmt 0)|
|I also would like to change some of my urls but am afraid |
It is unfortunate, but you should be if you change your urls the ad will probably stop running at least for 1 day (editorial review) then you don’t know if the change going to be good or all of your bids will shoot up also it might take weeks to get all of the former traffic back especially if you change the .tld extension. example.net to example.com now you are probably due for rough ride new domain - new history…
Now i think the best advice is if you make money, be very careful changing things around. Especially, consolidating keywords from one campaign/adgroup to another.
Reason - the minimum bids for the same keywords might be quite different if you move them to different adgroup/compaign.
Just my personal experience I added just one 1 ad and my min bids went down from 5c to 4c. But another person here on the board was complaining he added 1 ad and all of his bids went up form 5c to 10c. It is a killer if your breaking point is 9cents.
One of my ad groups spends $1500 a day and making profit and because of this uncertainty I am scarred to death too to add and test any more ads or add keywords in there. Even though I am quite sure if I am to continue to experiment eventually I would get better results and higher spends for google.
The quality score is now too sensitive and scares us to experiment with high volume well performing campaigns/ads/keywords.
Google is defiantly shooting adwords program in the foot here.
| 3:46 pm on Sep 12, 2007 (gmt 0)|
|I suffer from this. We have only removed stuff for months. The big fear is to get a quality hit to a major domain and lose existing campaigns. Also any new campaign is not as likely to settle in a manner like they did in the past so it is harder to absorb the initial high costs. |
Can not agree with you more!
Interesting observation here: in finance vertical where I bid $6.00 a click I don’t really care about initial bids that range for me from 5c to 30c but in another area like books etc 30c min bid is unbearable especially when it takes weeks for them bids to settle down. And personally I will not run a campaign at -80% ROI for 3 weeks enriching google even If I make small profit later – it is just very unfair practice here.
Seems to me Google loves you in high bids areas, and hates you if you are in the low bid range 3c-10c.
| 12:21 am on Sep 13, 2007 (gmt 0)|
Thanks very much for your continuing posts to this thread. I'll pass your comments along in, oh, about 29 hours.
Thus, still time for more. ;)
| 12:17 pm on Sep 13, 2007 (gmt 0)|
I too have a fear problem.
My campaigns cover many countries & when I originally set them up many years ago I just lumped them all together.
As time has passed & more competition has started running ads I find that I now need to split out my campaigns by country so I can control the bids in each one. As an example I need to bid higher in the UK & Europe than I do in the USA or Canada because of the currency differences.
I have been afraid to delete the UK & others from my campaigns & set up a completely new campaigns because I will lose all the CTR history &, perhaps, face large minimum bid changes.
Any suggestions would be appreciated.
| 3:14 pm on Sep 13, 2007 (gmt 0)|
|Fear is usually caused a lack of control or a feeling of lack of control. I certainly have that. |
Exactly! Lately Google Adwords has become more and more incontrollable. It is basically impossible for advertiser to understand what the various Google algorithms do, have done, will do....
You just add a bunch of keywords and adgroups, you hope that you get it right and pray. The logic is gone. Everything is very secretive!
Examples: I have the *same* keywords with different quality score in different campaigns (country targeting), both leading to the same landing page, some other keywords that do not show up at all (why?), new keywords that still need approval and so on.... yes I would like to re-orginize my campaigns around, but at what price?
If I delete a keyword, and move it to a new adgroup, do I loose my CTR, my history? Is CTR important anyway?
I tried to contact Google Adwords via e-mail and ask a few questions, I got template replies. Not useful.
SOOOOOO... Google did put mud in the water, and that may be good for Google, so the system cannot be tricked, but for us, advertisers, it is just lack of control and fear that if you change anything you may be doomed for life :-) (let's face it each advertiser is nothing compared to the total Google revenues)
Sorry for the long rant!
| 6:03 pm on Sep 13, 2007 (gmt 0)|
My understanding is that account wide performance does play a factor in Quality Score, that and a gazillion other micro-factors. That is the impression I always get when talking to Google Support anyway.
My main questions have always been, how are these factors weighted, and how far back is data taken into consideration, what is the "historical window" if you will.
The answer is always the same. That information is proprietary, if it were divulged then the gamers/black hats/bad people could manipulate the system. Fair enough I guess, but isn't there some kind of middle ground?
Seems to me a "rough" guideline as to the relative importance of factors would enable those trying to do it right more than enable those trying to game the system.
Fear is the bugaboo. Fear of competitors, fear of the system being corrupted. These fears of Google obviously spill over to the advertisers as evidenced by this thread.
I've often thought where Google may be the most vulnerable to competition is in the area of transparency. "Be Paranoid" seems to be the sub-text of "Do No Evil". Granted, the realities of competition and gaming the system are real. But does that mean we can't get a rough idea of how factors are weighed when calculating Quality Score? Even if they are just broken down into Primary and Secondary factors?
More to the immediate point.
If I am doing something really new or experimental, I may very well create a new account for that project. Not only because I do not know how much account wide performance is weighted in the Quality Score calculations, but also it gives me some added transparency when breaking down conversion and ROI numbers for my client.
[edited by: jatar_k at 7:58 pm (utc) on Sep. 13, 2007]
[edit reason] no sigs thanks [/edit]
| 3:10 am on Sep 14, 2007 (gmt 0)|
Many thanks for all of your comments - which I'll be passing along to the right folks, in this week's Advertiser Feedback Report, within the next few hours.
| 1:36 pm on Sep 14, 2007 (gmt 0)|
For what it is worth, I've had this fear in the past but I have found that generally making major changes like reorganizing campaigns, moving keywords to new adgroups, url, domain and ad text changes have not had any major negative side effects for me.
For example, I test new domain names from time to time and then if some keywords convert better with the new domain, I will just move them to new adgroups and send that traffic to the new domain while the other keywords will still point to the original domain.
Implementing new tracking URLs has worked fine for me and I've even started new campaigns by taking keywords from an existing campaign and reorganizing them in the new one..all with no ill effects.
A few "safety" precautions I use:
|1. Never delete anything, pause keywords, ads, campaigns etc if you are creating a new one. So if the changes for whatever reason don't have the effect you were hoping it's easy to revert back to the way things were. |
|2. Don't delete old ads when creating a new one. Create the new ad and let both ads run for a few days so you can limit any traffic "downtime" due to editorial reviews or syndication lag time. |
When everything is running, pause the original ad (if it is under performing the new ad)
|3. If moving keywords to a new adgroup or campaign, don't delete them from their original locations. |
I always set my maximum bid .01 higher for the keywords in their new locations. This seems to very effectively cause the keywords in the new location to show over their original locations since their max bid is higher.
Once you are happy that things are running smooth, pause the keywords in their original locations.
Also, in regards to fears of losing quality scores or keyword histories by moving them around, I've found that it takes very little time for the Adwords system to recognize that a high performing keyword in a new location has a good history and quality score in your account.
I may notice a drop in ad rank by a position or two for a day or so when a change is made but then things quickly get back to normal.
Deleting or moving a keyword does not eliminate it's history in your account but if you pair it with completely new ad text it will take a couple days for the system to reevaluate the quality score with that new ad so I always run the original ad with the new one I want to test and this seems to keep things running smooth.
I still get the jitters when I think of a better way to structure my account but so far I have had little to no problems in doing so.
| 4:28 pm on Sep 14, 2007 (gmt 0)|
jkwilson78 says it well. I echo the precautionary steps he outlines when making some changes. Like him, I too make what I would classify as moderate to substantial changes without too much adverse affect, maybe a day or two of lower position for the same cost.
But as I said before, if I am doing something radical (having warned or discussed with the client of course) I may very well fire a up a brand new account, especially if I am expecting poor initial CTR and/or ROI. The additional account insulates Quality Score and my clients conversion results.
Relative importance is probably fairly common sense. Account specific performance of a keyword would weigh more than public performance of the keyword., (yes, the word has a Quality Score component independent of any one account, this is more important initially when the word has no account/campaign/ad group specific history, so I understand.) KW performance weighs more than Ad group performance, Ad Group performance weighs more than Campaign performance, etc. But they all are considered in Quality Score at some level, apparently.
But we are left in the dark when it comes to the rough mix. I still think CTR is the most prominent factor in Quality Score, but is it 25%, 50%, 75%, dunno.
For instance would it corrupt the system if it were known that CTR accounted for approximately 75% of the Quality Score calculation, hypothetical of course. What damage would that info do?
But a legit manager would realize if he sees a drastic reduction in CTR as a result of lower bid/position he is also harming Quality Score, under this hypothetical anyway. May not make a difference if you can't generate profit at the higher bid, but at least you would be making a more informed decision.
The Great, OK, Poor grading system is better than nothing, but is there really only one level of Great for instance, I doubt it. My hunch is that given the same word, and a score of Great, a CTR of 10% for instance is going to result in a better Quality Score than a CTR of 3%, even though Grade for both may be "Great". I again get non-answers when I pursue this line of questioning.
I ramble, but I agree that this is an important topic.
Bottom line, Quality Score may be just transparent enough to be intimidating to those trying to do AdWords "right". Accounts that are just getting by, may go unimproved out of fear of losing what ground has already been gained.
Transparency, the new frontier.
| 9:52 pm on Sep 14, 2007 (gmt 0)|
yep, in the hope that my fears were unfounded, I changed a few ads last night. This am an unrelated ad is stopped do to trademark violation. The trademark is of course sort of like "widget". It is one word and means a common everyday thing which has nothing to do with what I am advertising. I have not had any issues in about 2 months and now less then 24 hours after tweeking 3 ads I have an issue on something that is really a non issue.