G A T is rubbish

[findtim]

i have an EMD which is an active online business for the last 2 years.

google adwords tool says the term has 170 broad/28 exacts/36 phrase p/month

but i have created my own tracking system and i'm getting 200 unique visits a week !

i've checked and 98% of inbound is from search not url typing as i have done NO advertising.

i'm happy as i am getting a nice return on the domain but would love to know why my stats "PROVEN" alter so differently from google

tim

 

[James]

Yeah I had a website ranking in Google US for like 4th position, and the term use to get me like 10,000+ uniques a month for the one specific term.

GAT said it gets 1,000 for the whole term, the thing is historically if the term is some what new it can take a long time to get updated.

GAT in the end of the day is a guide, the best type of data to use is proven traffic data/ PPC data.

 

[goldnugget]

I guess it depends over what sort of period G A T measures its average searches from. But on a brighter note it is encouraging to see that actual results could be much higher than what is posted from the tool rather than dramaticly under.

 

[eBranding.com.au]

GAKT numbers should be considered an approximate guide only.

I have some exact match domains which have GAKT numbers around xxx or less - but they get xxxx unique visits a month and make hundreds of dollars.

In terms of undeveloped domains, I've sold domains with less than xxx GAKT exact for hundreds, and I turned down an offer recently of $2,500 for a domain with 210 exact global monthly searches.

My point is, metrics aren't everything. Tools such as GAKT shouldn't be considered as anything other than a guide (albeit a pretty useful one).

 

[findtim]

i spent a bit of time testing my stats ( true ) against G A T and i have now concluded for myself G A T is a load of crap.

crap= based on emd, what i have seen is that many of my sites rank for multiple search terms so THUS other sites will as well, this is common basic knowledge, so creating a value of a domain on its emd exacts which i have found most people do is really missing the full picture.

i think we can over think the value of the name, a few things have happened in my buying and selling recently and unfortunately i have to say " snoopy's right "

the value of the domain is what someone will pay for it !

but some very well respected domainers i talked to at the xmas party ALL agreed that finding the end user increases the sale value no matter what the exacts were.

tim

 

[snoopy]

i have an EMD which is an active online business for the last 2 years.

google adwords tool says the term has 170 broad/28 exacts/36 phrase p/month

but i have created my own tracking system and i'm getting 200 unique visits a week !

i've checked and 98% of inbound is from search not url typing as i have done NO advertising.

i'm happy as i am getting a nice return on the domain but would love to know why my stats "PROVEN" alter so differently from google

tim

Your own tracking system? That doesn't sound very accurate. Personally haven't seen anything to suggest the adwords tool is particularly out.

 

[snoopy]

In terms of undeveloped domains, I've sold domains with less than xxx GAKT exact for hundreds, and I turned down an offer recently of $2,500 for a domain with 210 exact global monthly searches.

Should we then throw out the rule because of the exceptions? ...and typically it only gets thrown out when people already own a name and are looking to sell. When they are looking to buy the first thing they'll do is type the term into that tool.

People can see examples of names with weak metrics selling all day long at dnjournal. But the unless the CPC is very high these are still long odds sales, a few hundred sales a week out of 100 million names registered. They are not what someone can actually expect if they go to the market.

The real market is stuff like boat.net.au selling for $33 or sleepproblems.com.au getting $27. That is what your typical name with 200 searches a month is worth in a moderately commercial area.....$27. The difference to getting $2,500 is a 1:100 probability or increasing the probability by spending money/time marketing it.

Even then people would be extremely fortunate to be able to get the $27 drop price if they needed to sell & put it up for auction. They'd likely find no market.

 

[findtim]

Your own tracking system? That doesn't sound very accurate. Personally haven't seen anything to suggest the adwords tool is particularly out.

on what basis do you trash my system? geez you can be frustrating ! you have NO idea what i do, i think YOU think that told my whole soul on DNT, well you are wrong.

G A T is rubbish for making a decision on a domains worth, i can not believe i agreed with you and you still trash me

tim

 

[eBranding.com.au]

Should we then throw out the rule because of the exceptions?

You're drawing the wrong conclusions from what I said Snoopy, I think GAKT is a useful tool, the best out there for what it does - but I also think there's a lot more to domain values than just metrics. Most end users couldn't give a shit about metrics.

Have another read of my last point, I wasn't trashing GAKT at all, I was simply saying metrics aren't the only answer.

My point is, metrics aren't everything. Tools such as GAKT shouldn't be considered as anything other than a guide (albeit a pretty useful one).

 

[snoopy]

on what basis do you trash my system? geez you can be frustrating ! you have NO idea what i do, i think YOU think that told my whole soul on DNT, well you are wrong.

On the basis that it is likely no good, just run Google analytics if you want to track things. The fact that you are sure it isn't type ins doesn't inspire confidence because that it is the most basic thing that would be easy to see. What else are you missing? Probably a dozen other things.

The only way I know of fairly accurately testing the tool would be as follows. Run an ad on term and check that it is shows 100% of the time for a calendar month then compare what the tool shows for that particular calendar month, anything else is just guesswork. Even then you won't get search numbers, you'll get impression data.

 

[snoopy]

Most end users couldn't give a shit about metrics.

Unless it is a "brand" type domain I disagree. Endusers want traffic, and a keyword domain will potentially bring that, that is what it is all about. For domainers a keyword domain has enduser appeal and potentially a large number of buyers, the more popular the keyword and the more money in it (approximated by CPC) the higher than chance of a sale to an enduser.

 

[johno69]

i have an EMD which is an active online business for the last 2 years.

google adwords tool says the term has 170 broad/28 exacts/36 phrase p/month

but i have created my own tracking system and i'm getting 200 unique visits a week !

tim

I will add to this example.

GAKT says:

find poker - 73 local exact
findpoker - 22 local exact

Google Analytics say I got the follow traffic for the exact searches

find poker - 384 visits in the last month
findpoker - 146 visits in the last month

These are all organic searches for that exact term.

So i'm all confused about the numbers maybe someone can help explain.

 

[snoopy]

I will add to this example.

GAKT says:

find poker - 73 local exact
findpoker - 22 local exact

Google Analytics say I got the follow traffic for the exact searches

find poker - 384 visits in the last month
findpoker - 146 visits in the last month

These are all organic searches for that exact term.

So i'm all confused about the numbers maybe someone can help explain.

For a start there is no adwords data for November.

October shows 198 searches and it is increasing at a quite fast rate every month, December last year was 16 searches as a rough example of the change.

 

[johno69]

For a start there is no adwords data for November.

October shows 198 searches and it is increasing at a quite fast rate every month, December last year was 16 searches as a rough example of the change.

Ok I get it now. I'm creating the searches then basically, by people searching for my site.

Kind of backs up my theory that people search rather than type in.

 

[johno69]

October shows 198 searches and it is increasing at a quite fast rate every month, December last year was 16 searches as a rough example of the change.

Where do you find this data? I've only ever used the basic setting on that tool.

 

[James]

Just doing some research now, check this for example

Keyword 1: 2,875 visitors y-o-y (exact match GKT states 46 a month)
keyword 2: 2,861 visitors y-o-y (exact match GKT states 58 a month)

my advice take it as a guide, trust the PPC data

 

[snoopy]

Where do you find this data? I've only ever used the basic setting on that tool.

Need to download the report which will break it all down by month, show more detailed competition etc.

 

[johno69]

Need to download the report which will break it all down by month, show more detailed competition etc.

Thanks very much. Appreciate it.

 

[johno69]

it is increasing at a quite fast rate every month

I know this is a little old, but wanted to update it with some more data.

Local exact for term: Find Poker

Apr-12: 58
May-12: 73
Jun-12: 91
Jul-12: 110
Aug-12: 110
Sep-12: 140
Oct-12: 170
Nov-12: 170
Dec-12: 170
Jan-13: 210
Feb-13: 320
Mar-13: 480

But the initial figure shown when using the GAKT is still 170.

So it is best to drill down into the Local Search Trends to get a real idea of what is going on.

And I'm pretty happy to be creating such a trend of people looking for my site.

 

[snoopy]

I know this is a little old, but wanted to update it with some more data.

Local exact for term: Find Poker

Apr-12: 58
May-12: 73
Jun-12: 91
Jul-12: 110
Aug-12: 110
Sep-12: 140
Oct-12: 170
Nov-12: 170
Dec-12: 170
Jan-13: 210
Feb-13: 320
Mar-13: 480

But the initial figure shown when using the GAKT is still 170.

So it is best to drill down into the Local Search Trends to get a real idea of what is going on.

And I'm pretty happy to be creating such a trend of people looking for my site.

It shows the average, 2101/12 = 175