Our company manages nearly 50 AdWords accounts and has been using AdWords for several years. We currently have tens of thousands of keywords.
Historically, our strategy has always been to first launch just the exact match version of keywords as we find these to be our highest converting. We’ve never had problems immediately getting these words into a top bidding position at very reasonable prices – say anywhere from $.50 cpc to $1.20 cpc.
During the last few months, we have launched about five new campaigns. Every single one of these campaigns has experienced astronomical CPCs for our exact match keywords coming out of the gate. As an example, using the Google predictor tool, we find that Google predicts our keywords to be in a top bidding position of around $.70, but over the last few months, we are seeing new keywords with CPCs 3x to 6x higher than what the tools suggests. Also, these keywords are 4x to 5x higher than what we are paying for the same keywords over at Overture. Typically, we find Overture and Google to be fairly comparable. We are paying $4.00 to $5.00 CPCs for keywords we’re paying $.50 for at Overture. Interestingly, the Google predictor tools says that these $5.00 CPC keywords should be costing us $.60 in a first bidding position. We have moved the keyword positions down in positions to find drop offs, but the reduction in CPC has been minimal.
Again, to stress, this has only been happening on our newer campaigns, and is very, very consistent.
Recently, we even took a keyword out of a campaign that has been around for 2 or 3 years. We were paying around $.60 to have this keyword in 1st position. As soon as we moved the keyword to a brand new campaign, we had to pay around $3.00 to keep the word in first. I understand that when you move keywords to different campaigns, you’re starting all over, but this seems way, way out of the norm.
Interestingly, we’ve recently discovered that if we launch a new keyword in phrase or broad match only, the CPCs are much, much more reasonable for the equivalent bidding position of an exact match. The exact match is costing us 4x to 5x higher than the phrase and broad match terms for the same bidding position.
I understand that Google has recently changed how phrase match keywords are initially launched and that the phrase match keywords initially shows only in exact match until they reach a certain CTR threshold, but this change seems to have dramatically changed how new exact match keywords are treated.
My hunch is that Google is now assuming that our exact match keywords are going to have a certain CTR and is using this “assumed” CTR in determining our CPC. So if Google says we’re going to get a 5% CTR, and we actually get closer to 20%, we’re paying a premium to be in a higher bidding position. Is this correct? And if so, how long does this last?
SOMETHING IS DEFINITELY WRONG OR HAS DRAMATICALLY CHANGED. IT ALMOST SEEMS LIKE THERE IS A TECHNICAL GLITCH IN OUR ACCOUNT.
I have contacted our Google Rep and was told that “the field is more competitive” and “have we tried bidding in lower positions?”. This is absolutely bogus. If I go back into older campaigns and add new keywords, I have absolutely no problems getting those new keywords into a 1st or 2nd bidding position at a very reasonable price. Anything I try to do in a campaign started around the beginning of the year, I get absolutely creamed on everything I try to do.
I can not stress enough that what we’re paying for these new keywords in exact match is not even close to what the market value should be for these words. Something seems so very wrong.
Perhaps this is all a fluke, but I really don’t think so.
My questions are:
1)What do you think is going on?
2)Why is the keyword predictor so different than our actual results? I would understand variations for a handful of keywords, but we’re talking about 100 to 200 keywords we’ve recently added where the keyword predictor tool has been not even close to the actual CPC for what we end up paying. Not even in the ballpark.
3)Is it now better to launch new words in phrase or broad match instead of exact match? If so, why?
4)Can you provide some very detailed data on when a new phrase match keyword will begin to show exact and phrase match traffic?
I will do so, however, and also run it past a colleague or two - and see if I can offer any insight.
(I have to admit that when I have short periods to post, in between other task, I immediately gravitate towards the short question, and let the really long posts with multiple questions get pushed back repeatedly. Sorry about that!)
In the meantime, I certainly hope others will weigh-in as well, in spite of the thread's subject line!
AWA
Recently, we even took a keyword out of a campaign that has been around for 2 or 3 years. We were paying around $.60 to have this keyword in 1st position. As soon as we moved the keyword to a brand new campaign, we had to pay around $3.00 to keep the word in first.
This seems to be normal espcially when you take out an old keyword to a new start because it will be treated as a brand new keyword and will not enjoy anymore the CPC discount for your CTR histroy.
As you said you're paying $.60 right now for the keyword, actually this is the average CPC price caculated based on the historical figures. If you look at the recent click price, it may be much higher than the average one.
Interestingly, we've recently discovered that if we launch a new keyword in phrase or broad match only, the CPCs are much, much more reasonable for the equivalent bidding position of an exact match.
Again, the CPC you're looking at here is the average price for all the variations of a phrase or broad match keyword. The actual CPC for the exact match of this keyword may be much higher than the average CPC.
For example, you have WIDGET as your keyword and set it to broad (or phrase) match. Assume that all matches recieved only one click, and the CPC for the exact match is $0.6 and all other matches $.05,
Keyword -- CPC -- Clicks
Widget (exact match) -- $.60 -- 1
Blue widget -- $.05 -- 1
Red widget -- $.05 -- 1
Big widget yellow -- $.05 -- 1
In this case, the average CPC is $.75/4=$.1875. You can see that the average CPC is much much lower than the exact match price.
Of course this is over simplification but the idea is there.
Try to check, if it might be connected to the problem described here: [webmasterworld.com...]
To summarize, something seems to have changed where Google has allowed strong keywords with a strong history of high CTR, to become even more entrenched in a top bidding position, and conversely is making it more difficult for new keywords to reside in a top bidding position with a CPC comparable to others already in the top positions.
Anybody else seeing the same thing?
I read your post about a week ago and as a result, began a test earlier this week to determine the effect of having the campaign setting for United States versus all individual states.
I'll post our findings once I have enough data.
This morning, like a switch, our average CPCs took a giant tumble on our new campaigns. We are now paying something like 1/4th of what we have been paying for the last month. This is on multiple keywords in multiple campaigns. Competitors have not changed and no one has dropped out of the bidding.
For example, yesterday, I was paying $4.10 for a keyword I added a month ago. Today I am paying and average of $.60 CPC. During this time, I probably racked up about $15,000 of ad cost (perhaps overcharges) just on this keyword. This has happened on about 20 other high traffic keywords - all between last night and this morning.
We changed absolutely nothing between yesterday and today.
Anybody have any idea what the heck happened?
Seems like a glitch to me.
your experience.
So, you are not alone. Problem solved, you can say big Halelujah, if you look ahead. You will save thousands.
But if you look back ...
I am happy, that my competition is low, so my max bid is around 0.15 USD.
I'm now looking into this - and will post again as soon as I'm able. Many thanks for your patience. ;)
AWA
Any news on this topic?
I haven't heard back on this one yet, Tobias - but I'll certainly post again when I've learned more.
AWA
