Welcome to WebmasterWorld Guest from 18.232.124.77

Forum Moderators: werty

Message Too Old, No Replies

Tracking Overture clickthroughs

Using webTrends and tracking URL?

     
11:42 am on Jan 21, 2003 (gmt 0)

Junior Member

10+ Year Member

joined:Oct 31, 2001
posts:69
votes: 0


Hi,

I have used a tracking URL for a site which I am bidding on in Overture US, UK and Google AdWords. I've used 3 different URL's depending on the program.

However, acc. to Webtrends, these URL's have only sent approx 400 searches, whereas the combined traffic these programs sent to my website, acc. to the stats provided by Overture and Google, was 8,000! That means its logging only 5% of the actual referrals I am getting. I thought if I used a tracking URL, that referrals from partner sites would show up. But maybe they don't in WebTrends? Perhaps that is the discrepancy. So I suppose my question is Does WebTrends log referrals from Overture partner sites? I know this has probably been covered before but I did a search and couldn't find the required information.

Hope someone can help!

Thanks.

11:56 am on Jan 21, 2003 (gmt 0)

Full Member

10+ Year Member

joined:Mar 28, 2002
posts:211
votes: 0


i'm sure people will come up with a good answer to the discrepancy... cache and all that business?

but yes, webtrends shows partner sites...

3:50 pm on Jan 21, 2003 (gmt 0)

Junior Member

10+ Year Member

joined:Sept 23, 2001
posts:78
votes: 0


Just a recommendation...but there are Click-Thru scripts out there that you can buy to help you with this issue.
I had a Webtrends Sales rep tell me that its not as accurate as it could be...hence the part of the name "trends"!

Can you believe that?!?

Anyway, there are some perl and php scripts out there that you could use if you are just interested in a separate tracking mechanism to provide validation of what is being reported by the PPC's.

These scripts will also allow you to track multiple URLS.
I'll see if I can't find a url for ya.

3:08 pm on Jan 22, 2003 (gmt 0)

New User

10+ Year Member

joined:Jan 22, 2003
posts:2
votes: 0


there are better solutions than webtrends to track click advertising from keywords.

[TOS - no urls please]

[edited by: Mike_Mackin at 3:18 pm (utc) on Jan. 22, 2003]

3:18 pm on Jan 22, 2003 (gmt 0)

Senior Member

WebmasterWorld Senior Member 10+ Year Member

joined:Sept 7, 2000
posts:1720
votes: 0


Welcome whiskeythekid

I'm sure that you and/or SEGuru2 will be contacted if a member is interested in more information.

3:50 pm on Jan 22, 2003 (gmt 0)

Moderator

WebmasterWorld Administrator skibum is a WebmasterWorld Top Contributor of All Time 10+ Year Member

joined:Sept 20, 2000
posts:4472
votes: 1


IMO, WebTrends is not the thing to use for ad tracking. Like the WebTrends rep stated, it's good for seeing trends. If ya want exact numbers, or more exact numbers is more like it, go with something else.
2:21 am on Jan 25, 2003 (gmt 0)

New User

10+ Year Member

joined:Jan 8, 2003
posts:13
votes: 0


Unless there is something really tricky going on - like cloaking or a third party redirect or a funky firewall setup. I think that you probably have something setup wrong in WT itself. There shouldn't be such a huge discrepency. Cache's and proxies wouldn't explain a 20 fold problem.

More info about what version of Webtrends would help. If log file analysis, I would question whether you have the software configured to process your particular format of log files. If WebTrendsLive, then I would question whether you failed to put the snippet on some of your landing pages.

I would also double check your highest volume keyterm listings to make sure that your query string is in place.

I agree with the others about WebTrends not being the best tool. We use something called Conversion Ruler. It tracks pretty close to Overture's numbers (about 7% less) and gives ROI info. Feel free to sticky me about it.

8:36 pm on Jan 25, 2003 (gmt 0)

New User

10+ Year Member

joined:Dec 1, 2002
posts:18
votes: 0


Would it be a good/bad idea to do this:

Your desired landing page is widgets.html.
You make copies of this page and name them widgets-ov.html, widgets-aw.html, etc.
Use the source-specific page as your target url.

I make the assumption that any of the betters stats programs would be able to report accurate numbers of accesses per page.

4:58 am on Jan 26, 2003 (gmt 0)

New User

10+ Year Member

joined:Dec 1, 2002
posts:18
votes: 0


In the above response I left out I would put the new pages for the tracking in a new folder/directory and put instructions in robots.txt to not index the new folder. This would be to address the issue of "unwise" doorway" pages.
11:00 pm on Jan 26, 2003 (gmt 0)

Senior Member

WebmasterWorld Senior Member 10+ Year Member

joined:Jan 12, 2002
posts:1053
votes: 0


Your desired landing page is widgets.html.
You make copies of this page and name them widgets-ov.html, widgets-aw.html, etc.
Use the source-specific page as your target url.

The more you can segregate the PPC traffic from other sources the better, but, in doing so the site becomes unwieldy, you have to implement a significant robots.txt file to ensure that the hundreds of duplicated pages you use don't somehow find their way into the normal SERP's which often happens when you use site navigation as the skin for the landing page.

Industrial strength ad management tools are what we would recommend, there are plenty on the market, the figures are very accurate, you can measure as many campaigns as you like, you can account for the proxy effect, filter out your own test clicks, pick up valuable knowledge of search strings used (useful if trying to spot negative keywords or common words to exact match for higher CTR on Google Adwords), you can use it for banners, e-mail campaigns, newsgroup postings, so versatility is the key.

11:07 pm on Jan 26, 2003 (gmt 0)

Senior Member

joined:June 28, 2002
posts:851
votes: 0


Your desired landing page is widgets.html.
You make copies of this page and name them widgets-ov.html, widgets-aw.html, etc.
Use the source-specific page as your target url.
===========================================================

We started off doing this, ABSOLUTE NIGHTMARE.

as the site grows, and so do keywords etc etc, it got way out off control.

we rely on a simple script which can easily manage upto 10k clicks a day, life is so much easier now :)

Shak

11:07 am on Jan 27, 2003 (gmt 0)

Junior Member

10+ Year Member

joined:Oct 31, 2001
posts:69
votes: 0


Hi all,

Eric, to answer your questions, our client is using WebTrends Log Analyser to analyse several of their sites which are on the one server. They don't use cloaking or redirects but I'm not sure what their set up is regarding firewalls.

I think the software is configured okay, because the rest of the information seems pretty accurate. Surely if it wasn't configuredd to process their format of log files, then the reports would have no info in them? But they do contain lots of info about referrers etc.

The query string is in place also, because it is coming up, but just not for the numbers it should be. I think that part of the problem is that the clickthroughs from Overture's partner sites are not being picked up as being from Overture i.e. the tracking url is not being logged. Instead they are appearing as clicks from Yahoo and other partners. I'm not sure though, what if anything can be done in webtrends to overcome this. We may just need to get some more accurate ad tracking software.

Thanks for your replies!

ck

11:48 am on Jan 27, 2003 (gmt 0)

New User

10+ Year Member

joined:Jan 8, 2003
posts:13
votes: 0


I think you're right. If you're only counting the clicks from Overture then you would be missing a significant amount to the traffic. 20 fold difference would be possible.

But if you have the same sort of discrepency with Google, I would question the thesis. I get most of my Google Adwords traffic from Google itself.

I had presumed that you were somehow counting the referrals based on the query string, not the domain name. Is there no way to do that.

For what it's worth, I've never had a problem with partner sites failing to pass the query string . . . I can't imagine that they could possibly get away with failing to do so because so many clicks must be to DB driven shopping sites that need the query string to display the page.