Is AdWords a scam?

[ichthyous]

Hi there, I recently reactivated an adwords campaign that I had opened two years ago and closed after 6 months. I noticed then, and still notice now, that there is a fairly large difference in the number of clicks they are reporting and the number of actual hits to my site from my Ad. I wrote google and they gave a song and dance about web site logs, etc. I have two separate log analysis systems that are both very thorough. I also have my ad's URL hitting a domain that I bought and use only for this purpose, so i know that when I see a hit from that URL it's an AdWord hit. Still there is a big discrepancy between what I show and what they show. I am beginning to think that Google's Adwords are a bit of a scam...does anyone else have this experience?

[Tor]

We are managing more than 50 accounts and we have not experienced this at all up to now. There is no divergence between Googles numbers and our own numbers.

[Shak]

adwords is an absolute SCAM, please do NOT use it ever again.

Shak

[sem4u]

The less advertisers the better, eh Shak? ;)

AdWords is so not a scam. Why would so many companies be signed up to it if it was?

Log analysis programs and Google's stats are always likely to vary for a number of reasons. Have you tried counting the actual referrals from AdWords in your log files?

[ichthyous]

Yes I have. it's easy to do since they are all coming from the same URL...and there's a big discrepancy between what they report and what I see on my end.

[ichthyous]

It's also easy to count because I am running one ad with very strict keyword terms

[keywordguru]

To start what exactly are you using to track? Your raw server logs or third party trackers?

I would also say that depending on exactly where you have your ads showing you should be pretty accurate. Some trackers will only count an ip once rather than keep counting them as the same, so if your system is not up to date, it may be loosing some here. I would start small, and turn off both search network and content targetting, and run a few test campaigns with minimal dollars. Do this for a few days and see.

KG

[ichthyous]

I would start small, and turn off both search network and content targetting, and run a few test campaigns with minimal dollars.

This is already the case...i spend $2.00 per day for one ad, and both search network and content targetting are off.

To start what exactly are you using to track? Your raw server logs or third party trackers?

I am using AWstats and another raw log file analyzer

Thanks

[whoisgregg]

To start what exactly are you using to track? Your raw server logs or third party trackers?

I am using AWstats and another raw log file analyzer

Do they show the exact same numbers? I've used Webalizer, analog, and summary for log analysis and can get wildly different numbers with the same raw logs. (Essentially attributable to how well each one plays with corrupt log file lines.)

[anallawalla]

i spend $2.00 per day for one ad

We spend in the four digits each day so we have far larger numbers to compare. We use Google's own conversion tracking, our own web logs, an external bid manager tool, and a fraud-checking tool, so we have many numbers to compare. They are never identical, but the variance is low enough to be ignored most of the time. Google also refunds what it considers invalid clicks almost every other month, although I have no way to tell if it is accurate.

[Paris]

I too have never had a problem with AdWords not closely matching the logs.

Are you sure both days are the same? Google's stats are done in West Coast time and if your web log isn't lined up with that it's probably not going to match.

[ichthyous]

What is the name of your fraud checking tool? Thanks

[The Grizzler]

We have just installed some scripts on all our pages on our site that track the user live. I have spent 12 hours today glued to the screen watching users going from page to page, checking referrers etc.

Allowing for cache servers etc, I cannot believe what I am seeing. I reckon at least 40% of clicks on our Ads are fraudulent.

So even if the log files match with the Google figures, I am still paying for fraudulent clicks anyway.

We need to trust Google implicitly to relax. Yes there will be differences in logs from Google figures. But how do we quantify the accuracy of this?

[anallawalla]

Fraud checking tool: Whosclickingwho

Disclaimer: I haven't had it long enough to recommend it for that purpose. It has some side benefits, such as seeing the site where an AdSense ad was clicked or the keyword used. It is quite a different view than one from our regular log analysers because you see the keyword next to the specific URL, whereas the log analysers give us a sterile list of keywords used.