Google Exacts versus Actuals - Examples

[archibel]

I suspect most of us assume that Google Exacts always overstate the true search numbers and I think it would be interesting to share some evidence.

For example, the following three websites all rank number 1 in Google for the main search term - I’ve shown the number of Google Exacts for this term and the number of actual visitors who arrived at these sites by entering the phrase in October (none of these sites have seasonal variation and I'm using AWStats).

Website 1 = 2900 exacts (404 actual)
Website 2 = 5400 exacts (702 actual)
Website 3 = 3600 exacts (827 actual)

It would be great if people with high ranking websites would be happy to share some additional data.

 

[boxfish]

Your title and description can impact your click through rate, so just because your site only gets so many visits, doesn't mean that's all that's on offer.

A while ago I regged a domain with an estimated 18k exacts a month

I set up an adwords campaign to display only for an exact match, ran it for a few days and then averaged the impressions across the month.

Worked out at just over 23k a month

 

[arv]

There must be ads above your site which may get the clicks. Check the webmaster tools and see how much impressions do those keywords actually get on the web. Sometimes, those phrases may also be searched in google images where you may not have received any clicks.

 

[archibel]

Yes, that's a fair point Boxfish - but I think my titles/descriptions are as good as they could be. E.g. try a Google search for 'false nails' (Website 3)

 

[donreeco]

Thanks for sharing.

Do the serps looks the same for all three sites or are there obvious differences, such as the number of ads above the first position, local listing etc, which would obviously reduce the CTR of a first place position?

 

[archibel]

Donreeco - maximum of 3 sponsored ads above the organic position (no images, maps etc)

 

[Blossom]

It's impossible to make a fair assessment without holding all the top ten places (absolute minimum top 3) plus all the ads. Realistically that's not going to happen.

Boxfish's way makes more sense, but to get the most accurate figures you'd have to run a campaign for a really long time to account for seasonal differences.

 

[mally]

not sure if because its a .com as most uk users including me would look for a .co.uk to click on, as I would think that site is aimed more towards the US

 

[newguy]

I think you're right when you state that exacts are not always accurate. However, in the examples you give, are you assuming that the number of exacts and actuals will match? Wouldn't that mean that you're expecting everyone who types in that term to visit your site, and not a percentage of them? Maybe I'm missing something, but your actuals and about 20% of exacts, which isn't that bad really.

 

[archibel]

Thanks Newguy - I appreciate the sponsored results will soak up a large percentage of searches...I was just hoping to understand whether the % click-through I'm seeing across these three sites (and others) is reasonable. But perhaps there are too many variables to understand whether the Exacts numbers are too high.

 

[newguy]

Thanks Newguy - I appreciate the sponsored results will soak up a large percentage of searches...I was just hoping to understand whether the % click-through I'm seeing across these three sites (and others) is reasonable. But perhaps there are too many variables to understand whether the Exacts numbers are too high.

Yes, I think it's a tricky one to gauge. It definitely does feel 'out' for some exacts though.

 

[vizzo]

I had a site at one time that was getting more google visitors per month for a particular search term than the exacts in the GKT - so i guess it cuts both ways!

 

[djbw]

I could have sworn I posted in this thread already...

My best ranking site is first for the plural (480 local exacts) and sixth for the singular (590 local exacts), and I had 340 visitors last month.

 

[stevieboy1101]

org.uk

i think it raises the question of the effectiveness of .org.uk and .me.uk domains in terms of traffic leakage and the value of these domains when you come to buy them.

 

[David]

If you're referring to Google's keyword tool surely the key thing here is impressions - not site visits?

i.e. we should be comparing impressions for a ranage of exact match keywords (with 100% impression share) to what Google estimates?

I've done this using my Adwords account for non-seasonal keywords and it's looking like Adwords generally underestimates volume...