WebTrends cookie rejection

Connection between cookie rejection and spiders

[dejamant]

Hello,

I'm using WebTrends 8.5 and report Marketing > Visitors > New vs. Returning Visitor shows that 73% of visitors reject first-party cookie. On the other hand, report Site Design > Browsers & Systems > Platforms shows that there is 31% of spiders.

My question is, can I make a some kind of connection between these two reports, and find somehow does this 30% of spiders make 70% of cookie rejections?

Thanks in advance.

[cgrantski]

To an extent. A lot of spiders don't accept cookies.

At the same time, you are probably getting quite a few spiders that are getting past WebTrends' spider definitions. If you look at your Top Visitors report you can probably identify some of them just by looking.

Are you analyzing server logs or using javascript data collection? The interpretation of the "no cookies" table is very different for each.

[dejamant]

Indeed, you are right. I have various spiders as Top Visitors. For now, WebTrends is analyzing server logs, but I'm will soon switch it to JS tagging and SDC. Will this improve spider filtration?

Thanks for the reply.

[cgrantski]

Yes. the javascript method of data collection avoids a lot of spider traffic. There is a long recent discussion on this (in the last week or two) here on WWW.

You can also stay with server logs and create filters to get rid of the ones you are seeing and can identify.

If you are using server logs, the "no cookies" number will include any visitors who have never been to the site before (i.e. don't already have your cookie) and who only stayed for one page (i.e. were never able to demonstrate to WebTrends that they could accept cookies, and WebTrends just assumes that they don't). WebTrends really should call that row in the report "cookie status unknown" instead of "not accepting cookies", at least for people using server logs.

If you use the javascript method of data collection, the WebTrends tag does a very smart thing - it executes a cookie-acceptance test on the very first hit and records whether they accept cookies, even if the visitor only looks at one page. It's completely different and much better than the server log method. So, with SDC data collection, you have a very accurate idea of cookie rejection rates.

Hope this helps. There is a good WebTrends user forum out there, reachable from the WebTrends site. And there's another WebTrends help blog called the WebTrends Outsider. None of them have the breadth of what's here on WoW, but they are good resources.

[dejamant]

Thanks again for the reply and the resources.
It's very helpful.