Goto vs Webtrends

Why are statistics so far off?

[legster]

Justing using the example of one keyword phrase here. Goto says that I get on average around 50 clicks per day for this one keyphrase. If I check in Webtrends(we have the most expensive, nicest version) it says we only had 5 visitors for the same phrase. Does this mean that my visitors are clicking the search listing ten times? I know I have heard a lot of talk about some inaccuracies with Webtrends, but this is a huge difference!

If the problem is with Goto that is a whole other story and a major problem! I do understand that there are competitors who will keep trying to click listings, and that Goto can't stop all of them. I have noticed this wide range though seems to be consistent across all domains, and keywords. So I at least doubt that competitors are that vigilant in clicking. If they are, than that would me around 90% of my Goto clickthroughs are fraudulent and that would be bad!

What do you think?

[seth_wilde]

It's probably because most of your goto traffic doesn't come from goto, it comes from goto's parteners. You should set up a tracking page that can only be accessed via goto.

[asuprun]

Hi Seth,
By doing that you will not avoid being clicked from GoTo partner sites anyway. However, is it possible to avoid click charges from other GoTo partner sites in any legitimate way?

[Mike_Mackin]

90% of the traffic will come from the partner portals.
[goto.com...]

[tedster]

The easiest way to track clickthrough that I've found, is to follow GoTo's suggestion -- register an URL with GoTo that looks like this:

www.site.com/page.htm?src=goto

We started doing this about 9 months ago with one client's campaign and the tracking numbers we get are really close to what GoTo charges. This tracking method avoids creating pages with duplicate content -- that's a maintenance plus and a search engine plus.

The partner sites are the point of the campaign -- this client runs only 5% of traffic from GoTo.com itself. The GoTo clicks also have a lower conversion rate (an effect of competition checking things out, I'd imagine).