Accuracy of Adwords click data

Hitbox showing 40% less traffic

[mhkatz]

I have an Adwords campaign with all of my ads going to a specific landing page. When I look at the statistics for that landing page in Hitbox, it shows about 40% less visits (as entry pages) than what my Adwords account shows for the same period. Can anyone explain this discrepancy?

[Sasquatch]

Does hitbox still collect all their data through images linked to their site? Or do they now have a log analyser?

The reason I ask is that I know hitbox is blocked by almost every piece of privacy software out there. If I go to your site, no requests will be sent to any hitbox URL and I will not accept any cookies from them.

On the other hand, if you are analysing your logs, you might have a reason to complain.

[shady]

Thats very interesting, Sasquatch. I have written my own hitcounter, as I wanted to monitor traffic through a source filter so I could easily view the traffic from each of the PPCs and also free clicks.

I use a simple javascript call:

xxx = new Image(1,1);
xxx.src = "http://xxx.xxx.xxx.xxx/params etc.....;

Will this be blocked by certain browser configurations (other than js being disabled!).

If so, is there a better way?

I've found results to be more accurate than other well known hit analysers. One reason may be that I am using the IP rather than the domain name, which may make it more reliable.

[Sasquatch]

It should not be a problem as long as you keep it on your own domain. I think there are some of the programs that will block the web bugs though.

Since you had to write your own hit counter using some form of cgi, why don't you just write your target pages in whatever language you are using?

In PHP it is incredibly easy to get to all the cgi and server variables. For example, evey hit by googlebot I write to a mysql database for much easier access than looking through the logs.

[mhkatz]

Sasquatch,

Hitbox, or at least the version I am using, does not analyze the log file.

I do not understand how privacy software would effect the counting of referrals from Google, since Google is the referrer, not the individual user.

[Brad]

Hitbox misses a lot of visitors. If you compare actual server logs to Hitbox stats you will see they can miss +33% easily. All of these remotely hosted stats trackers miss hits not just Hitbox. They are better than nothing if you do not have access to server logs.

[progen]

I found that a lot of the time the code fails to load or takes too long and the visitor changes pages.

From experience I found hitbox especially bad at missing visitors although it does provide great stats for the one's it logs.

[shady]

Hi Sasquatch

Actually, the hitcounter is on a different domain/server.
I run many sites and like to have a central point when I can easily monitor traffic.

I guess it works in the same way as most commercial hit counters.

Is the block that you speak of, one which disables an html page from accessing components from a different domain/ip?

[Sasquatch]

mhkatz,

Hitbox warehouses data on users and can track their surfing habits across various websites. Many people find this to be offensive behavior and they block it.

To get accurate numbers you need to track data from your real logs. And even then, some people block their referers when they surf, and if your site is slow, some people may give up on your site before they get there from google.

What about if your website is down while someone is trying to follow your ad? Google will count the click, but you will never see it, even in your logs.