Post Reply
Posts: 1097
Joined: Fri Jun 22, 2007 12:40 am inaccurate?

Post by wehaveitall » Mon Jun 29, 2009 4:30 am

I contacted a website called about advertising I checked their website on Compete, and found that it got around 15,000 unique visits a month. The webmaster sent me a picture of Google Analytics showing several thousand visits a day. I had also noticed that on Google analytics had 500 unique visitors more than what listed.

The thing about the image sent, however, is that the website title is cut off, and only the numbers are shown.

Is way off, or am I being lied to?

Posts: 6987
Joined: Sat Oct 07, 2006 11:43 am

Re: inaccurate?

Post by Sami » Mon Jun 29, 2009 6:36 am

Compete and similar sites can only estimate traffic and they are pretty much always inaccurate. I personally know websites that receive 10x more traffic than Compete says.

Posts: 4906
Joined: Thu Jan 03, 2008 1:46 pm

The term "visitor" is also fuzzy

Post by phantazm » Mon Jun 29, 2009 1:44 pm

"Visitor" is an example of something that everybody knows. But there is no standard definition of when a visit should be counted. Consider these examples:

If I visit your site, I'm a visitor (and counted as such). If I revisit, how long do I have to wait to be counter as another visitor? 1 minute, ½ hour, 1 hour, ½day, whole day, month, year, or what. Standards differ wildly here...

If I visit your site, from my own pc, from a public library, from a friends pc, etc - am I then counted as 1, or many..?

If I visit your site for the first time, then I'm unique. If I revisit 10 years later, I'm unique again. How about next year, month, day, hour? When am I unique again..?

If I visit your site only once, but copies an url of an image and use it on my own site. Are my visitors then also partly yours..? (Some years I had a statistic that behaved just like that: any "hit" was counted as another "visitor". Stats were impressive, but also became absurd, as there were more visitors than pageviews. But how can one visit a site without viewing at least one page, right..?)

I prefer checking number of 'visitors' and how long they stay.
What did they look at? And where did they come from?

Defining the exact number of visitors is just as impossible as measuring the length of a coastline (the length depends on the size of your ruler (=fractality))...

Posts: 1097
Joined: Fri Jun 22, 2007 12:40 am

What's the best way

Post by wehaveitall » Mon Jun 29, 2009 1:45 pm

How can I verify this guy is telling the truth?

We rate the websites, the WOT staff creates, and advertises the add-on, and together, we make it all happen.

Posts: 6987
Joined: Sat Oct 07, 2006 11:43 am

Re: What's the best way

Post by Sami » Mon Jun 29, 2009 2:17 pm

How can I verify this guy is telling the truth?

I think it's possible for him to allow you access to the site's Google Analytics dashboard. It might also be possible for him to have Google send you an email containing statistics for a certain time period. Then again, a difference of 500 visitors isn't that significant...

Posts: 22
Joined: Sun Jan 04, 2009 9:56 pm

I'll give you an example of

Post by app103 » Sun Jul 05, 2009 3:43 pm

I'll give you an example of how inaccurate 3rd party analytics really are. is one of my sites. I have Google Analytics code on every page, as well as having Awstats, server-side.

If 3rd party services were accurate, then the results should be pretty much the same with what I have reported on my server, right?

Well, it isn't.

Here is what Google reported for May 2009:

1,457 unique visitors
1,673 visits
3,230 page views

Here is what Compete says about that same period of time:

1,733 unique visitors
1,894 visits
(no page view data unless I pay them)

Awstats running on my server:

4691 unique visitors
5954 visits
15,884 page views

I don't know where Compete gets it's data from. I do know where Google is getting it.

But I also know that anyone that is using a noscript plugin or anything else that blocks the loading of their script and the setting of their cookies won't be in the statistics, which is why it is much lower than what is reported on my server.

I would trust the stats from Awstats more, since it isn't based on info collected through inserted javascript and cookies, and it is kind of hard to block the server from collecting the data without blocking yourself from connecting to my site, completely.

I believe Awstats is using IP addresses as the way to identify a visitor. If the same person comes back but with a different IP address, they would be considered a new visitor. But Awstats page views data is very accurate, as it counts every time the page is requested from the server.

If you want transparency in advertising, I'd go to Project Wonderful and find sites to advertise on through them. You can see based on stats collected from the ad widget, how many unique visitors are being displayed the ads, and how many page views, and top referring sites, currently running ads and how much they paid to be there, and the bidding history. Project Wonderful makes this info all public for any site carrying their ads.

Here is the public data for my site, so you you can get an idea of how it works:

And it's pay-per-day advertising, not per click. And you bid on it, like ebay, and for as long as you are highest bidder, your ad will run.

It's also possible to advertise for free on lower traffic sites that have empty ad slots. For example, I got a years worth of advertising on this one guy's site for $1.34, by bidding $0.01 per day, and getting it for free when I was the only bidder on the slot.

The info they give is more important than what you would get from Compete, Google Analytics, or any other source, since it is data that pertains to the actual widget your ad would be appearing in. Who cares how much traffic a site has if visitors are running ad blockers and will never see your ad? I'd rather have the data related to who is going to actually see my ads.

Post Reply

Who is online

Users browsing this forum: No registered users and 3 guests