Google Keyword Tool Analysis - Huge Anomalies

Not open for further replies.
Has anyone else experienced huge differences in Keyword results?
Example, I searched back in March for a few Terms, found a nice double combo with a good strong Local (Local = Ireland, Lang = Eng etc etc) Monthly of 720. It had a nice even spread across the 12 month range.
So we purchased the .ie domain based on this combo.
Here we are now, site almost built and ready to go live and I see the Monthly result with a measly 22!!!
Its not even high enough for me to see if some months are better than others as sadly it's not even displaying monthly Search Trends for that combo.
I had even downloaded the raw CSV back in March so I know I'm not going made, there it is almost top of the list.
Anyone experienced anything like this before?
I guess if I had to go back to IEDR and get yet another discretionary name for the client it wouldn't be the hardest thing in the world, but how can I be certain those listed terms wont fall of the radar too?
I know there's bound to be monthly fluctuation of course, but for a relatively non seasonal product 720 to 22 is just beyond possible no?


Staff member
That kind of drop sounds dramatic, but could it be due to media coverage or some news event?


New Member
It's not likely that it dropped, it's more likely that the first number was a guess and then your site's CTR-Impressions helped Google correct it. If there's a low search volume, Google borrows the value of any phrases it thinks are similar. It's also able to suggest that if there are 3,000 searches for "water taps" and "4,000" for "water drums" then "tap water drums" could have 7,000.

There's a number of reasons for this:

1. It's a computerised aggregator
2. Numbers are factored, rounded and calculated - so the number of searches in a day are multiplied by 30 (for example) - giving lots of large numbers like 2,220 (not 2,221)

But the biggest reason is AdWords Matching. You see if you sponsor a word like (water) but don't give any matching (e.g. "water" or [water]) then, counter-intuitively as it sounds, - H2O matches it. And so can liquid, fountain and pottable. Broad matching doesn't require the word(s) to even match.

So if Adwords says there are 70,000 searches for (water) and 7,000 for "water" and 700 for [water] - its the last one that is the most accurate.


New Member

If you are number 1 for "water" and your site is called "" and you have sitelinks, your average best Click-through is about 49%. It's really about 37%. Check your CTR for actual searches -v- clicks.
Thanks for a super response as ever Link8r! As usual this leaves me scratching my head somewhat agasp at how little I really know, or better still how much I have to learn. :)

Michael Brookes

New Member
I've come across discrepancies in the keyword tool before, link8r covered pretty much everything but I thought I would add, selecting "exact match" on the left hand side gives far more accurate search volumes for selecting a domain name to match a keyword.
Not open for further replies.
Award-winning Mac antivirus and Internet security software