VKI Studios is now Cardinal Path! www.CardinalPath.com
Learn more about Cardinal Path

Display the Source of a 404 Error in Google Analytics

While falling asleep about five minutes ago (looking at the longtail keywords for our blog) I was struck with a pretty awesome idea. If you remember Fixing Your 404's with Google Analytics, I mentioned Matt Cutts solution, using Google Webmaster Tools to track where 404's are coming from. Well, using a cool combination of a couple of Google Analytics tricks, I've realzied that you can duplicate that same functionality in Google Analytics using a combination of the "full referrer" filter, and error tracking. This is currently just an idea, and I'll have to test it and play around with it to get any bugs ironed out, but this should provide a way for you to pinpoint links to non-existent pages on your site.

Here's how you do it:

1. Tracking your 404's

The first part of this involves a previous blog post of ours. By playing your GATC code on your 404 error pages and then creating a quick filter, you can track visits to error pages. Here's the filter:

Filter Type: Advanced
Field A -> Extract A | Page Title | (Error [0-9][0-9][0-9].+)
Field B -> Extract B | Request URI | (.*)
Output To -> Constructor | Request URI | $A1$B1
Field A Required: Yes
Field B Required: Yes
Override Output Field: Yes
Case Sensitive: No

This takes the title of the page (so error (three numbers) and whatever), then also takes the URI that the user requested, then combines them and outputs them, overwriting the requested URI string.

Using this, you can now see what pages are causing 404's by using a data filter (at the bottom of a report) to and filtering by titles containing "404".

2. Displaying Full Referrers

This is just another filter.

Custom Filter
Advanced
Field A -> Extract A: Referral (.*)
Field B -> Extract B:
Output To -> Constructor: User Defined $A1
Field A Required: Y
Field B Required: N
Override Output Field: Y
Case Sensitive: N

This takes the referral string and outputs it (raw) into the user defined value.

3. Combine the two

Now you just have to build the report.

  1. Go into top content
  2. Filter by "404"
  3. In the secondary drop down box, select "user defined value".

Now since I just thought this up, and since it takes time for GA filters to propagate, I can't tell if this will work flawlessly, but I can see no immediate reason why it shouldn't. I'll test it for a week and update this post with my findings.

Comments (Comment Moderation is enabled. Your comment will not appear until approved.)
Sounds like a good idea! I'm giving it a try and I'll let you know how it goes.
# Posted By Stephane Hamel | 10/28/10 5:50 AM
Seems like this work perfectly. I will also reference this in my blog.

Step 3; shouldn’t 3 come before 2 (Filter by "404")? ;-)

Regards,
Oluf
# Posted By Oluf Haugen | 11/11/10 1:34 PM
.