Has adword tools gone wonky !!!! i'm getting very odd results since last night.

[justwondering]

Hi , i know i started another thread but that one was explained...

I'm not logged in and i've tried this on 3 computers....in case its something on one of my computers ( i've got six at home....pretty sad really lol )

Basically , i was checking out adword exacts from around 9pm last night.....found quite a few interestings exacts...anyway i decided to double check various things at around 11pm , i was getting back totally different results ....i double checked...and even used "model kits" as a bench mark .....

this is currently showing 4400 exacts worldwide and 1300 local exacts.!!!!

is anyone else having the same issue...if possible could you input :- model kits , into adword tools and see what results you recieve..

previously this was showing around 301,000 for uk local exacts

has google adwords gone wonky ?

 

[Edwin]

Has it updated from July to August data, perhaps?

 

[retired_member13]

Has it updated from July to August data, perhaps?

Don't think it can be that - the new KWT shows an average of the last 12 months, rather than just last month's figures.

I'm seeing the same numbers as the OP btw.

 

[justwondering]

massive difference in results from last evening....i also tried this on my 3g connection , just in case it was my i.p !

same results...dramatic difference especially :- model kits , from 301,000 to 1300 !

 

[retired_member21]

Me too - 135k exacts to 18.1k on one of mine

 

[springer]

Probably some technical fault. The previous interface seems to be working OK.

 

[justwondering]

hopefully not a correction.....lets hope adword tools has been getting it right and the current figures are incorrect.

 

[aZooZa]

I've seen the figure 301,000 so many times.

 

[Darren]

I was getting this error when logged in, was fixed by logging out.

Now its the same results whether logged in or out.

Darren.

 

[justwondering]

what's the amount on the figure gold frog ? are you seeing 301,000 for model kits or the much lower figure ?

 

[Darren]

With the new GKWT, 1,300

Darren.

 

[Brassneck]

Doesn't 301,000 seem unrealistically high? Would have thought a lot of searches would be 3 word searches for this - eg airfix model kits.

 

[justwondering]

Doesn't 301,000 seem unrealistically high? Would have thought a lot of searches would be 3 word searches for this - eg airfix model kits.

indeed it is ! but what a huge drop , how many have your good exact domains dropped by ?

 

[namealot]

It is often not accurate but its better than many of them... It wasn’t designed for keyword exact search volume it was a tool designed for advertisers interested in Ad words …? What it really shows is the number of times your ad would appear for that keyword and that's not the same as actual searches…?

If you achieve top place on g you will see you’ll be lucky to get 5 -10% of what it shows you should be getting

 

[retired_member21]

If you achieve top place on g you will see you’ll be lucky to get 5 -10% of what it shows you should be getting

That's a matter of opinion and is dependent on a number of factors including:-
- # of PPC adverts
- sitelinks
- double listing (double whammy )
- quality of meta information

Around Jan/Feb this year I was running between 20-40% exacts at a #1 position without sitelinks for a competitive phrase.

I wouldn't necessarily go on 5-10% - I'd say #1 spot is more likely to be around 25% on average (in my opinion!)

 

[namealot]

Yes I know g will trigger ads based on past searches, personal search , geographic locations etc.

Some cynical people might go as far to say its trying to stop the long tail growth, to increase bids for head terms and generate more revenue...

Just I think alot of people don't understand its not keyword true but adword so take it with a pinch of salt and not that its gospel

 

[justwondering]

its a common point of reference "exacts" when stating the so called potential of a domain name , one i have hand regged recently have dropped from 22k local exacts to 140 local exacts....thats a massive difference .....but i'm also wondering....based on some very popular terms such as "computer repair" it states local exacts at 3600 ! i'd take a stab in the dark and say that does not sound right for the whole of the U.K !

 

[Brassneck]

its a common point of reference "exacts" when stating the so called potential of a domain name , one i have hand regged recently have dropped from 22k local exacts to 140 local exacts....thats a massive difference .....but i'm also wondering....based on some very popular terms such as "computer repair" it states local exacts at 3600 ! i'd take a stab in the dark and say that does not sound right for the whole of the U.K !

Using legacy tool I get 40,500 local exacts for computer repair - that does sound about right. I don't ever use the new tool.

Stephen.

 

[JMOT]

Don't just look at the exacts...

Look at how much a lead or sale is worth in commission...

 

[retired_member13]

Well, I think that one of two things is happening here:

Either:

1) It's an error or

2) G has redefined the criteria for what counts as an exact match.

Historically, as far as I'm aware G has included multiple page views in SERPs as separate exacts matches (ie viewing the second page of results or returning to the first page [uncached] after visiting a site adds +1 to the exact match total). Also up until now the figures have been very susceptible to being manipulated by auto queries, which has resulted in some very odd readings for obscure terms. Obviously there are likely to be other factors, such as stats from Google's ad partners, parking pages and so on.

Removing one or more of these factors would result in lower, or much lower readings.

I've got quite a few very high exact match sites sat at #1 in G for terms, and often the traffic they get bears no relation to a traffic estimate based on user eye scan / click rate studies and the quoted exact match figures, so assuming that this is not a system wide error it will be interesting to see if the reality matchs up better to estimates based on the readings that are now being given.

As has been said by many members both here and on other domain forums, valuing undeveloped names solely on exact match without adding a good dollop of experience and common sense was always going to be a risky strategy. Regardless of the size of the readings given I do think that the tool has a value in terms of adding metrics to the buying equation for selecting target names.