Regular Google vs. Adwords Clicks

[web_young]

I am tracking what people are searching for to find my site. I was wondering if there is a way to distinguish between referrals from the regular Google SERPs and Adwords. I want to be able to figure out which referrals are coming from regular SERPs and which are coming from Adwords.

[Macro]

bump

I'd like to know this as well if anyone has the answer.

[dmorison]

Not easily; the only thing you can do for sure is tag an identifier onto the URL (the classic being?source=adwords); but this of course makes a mess of the address bar...

A referrer is a referrer and cannot therefore be used to distinquish between a free SERP click and an Adwords click. Note that sometimes both SERP and Adwords clicks are channeled through various "qulity control" monitoring scripts at Google.

[Receptional]

Adding?source=adwords is a good plan. We use a javascript tracking system that then filters out Google Natural from Google Adwords based on this. It seems to work.

Of course in THEORY (but I bet not in practice) you could look at your tracking to see the TOTAL google referrals, then go into Adwords and see the number you have paid for, making the difference (presumably) Google natural. Now I can see many problems with this - but you might look at the numbers and see if they sound sensible.

Also, assuming you can track by search phrase, you will see some traffic using phrases that you never bid on.

[percentages]

Make the AdWords URL unique.

Currently I see my clients getting 95.7% of referrals from true SERPs and the other 4.3% of Google referrals from AdWords......given those figures I have to wonder if it is actually worth measuring;)

[karembeu]

What about a situation where you have a LOWER number of listings in your log files than adwords is giving?

Yesterday adwords showed a more than 200 click discrepancy in their favour, logfiles (using WebTrends) showed just shy of 600 clicks from Google, AdWords showed 800 clicks?

Any ideas?

[Receptional]

What log file analysis system are you using?

[karembeu]

webtrends 7.

[Receptional]

Then look at the caching issue at [webmasterworld.com...]

Dixon.

[chewy]

I am seeing what Karembeu is seeing as well.

I use both WT 7 and WT Live.

I don't get much satisfaction from reflecting on the cache issue, or what Google says, or what NetIQ says. And believe me, I have looked, tweaked the numbers, picked up the phone, stickied people here and scratched my head long and hard about this.

I need some input here from you all.

My theory is that there is an increasing number of browsers and or proxy servers that are filtering or failing to pass the referral string from Google.

Conspiracy theorists might say that Google and others have a vested interest in dampening the statistical validity of these numbers -- but I don't go there.

My view is this is just a conspiracy of inadvertent simplicity.

Anyone else got any theories or ways to test this or that hypothesis?

[sun818]

In my AdWords campaigns, I put append the URL with a unique identifier such as?adw. This distinguishes if the click comes from a Google search (left column) or from Google AdWords (right column). I think its also helpful to track what page your visitor lands on since you can potentially have two web pages (the main, and an indented one) for a search term.