My first infographic, in the form of a political joke. Enjoy! And in case you missed it, that is Julia Roberts from Pretty Woman..
Posting a sample page of a blog with a comment spam issue. Just an example. I’m sure someone out there will probably find this useful. I wonder if all those spammed sites are properly sanitizing their data inputs. And yes, this is one of my blogs. I lost interest about 15 minutes after registering the domain.
Maybe someone will find this useful. There seems to be a lot of research on blogs these days..
Testing Google Analytics filters is pretty damn hard. It’s even harder if you don’t realize that the filters take forever to activate! How long do Google Analytics filters take? In my experience, around an hour or more (it’s not always consistent).
That’s right – even though you are watching the GA data update itself, roughly every 10 minutes or so, do NOT assume those filters are acting upon the data. I found this out the hard way. After tediously testing Google Analytics filters for the better part of a day, they FINALLY started to weed out the traffic noise I was targeting. I ran a few more tests and waited for the traffic to percolate through. Working fine so I left the office for the day. I returned tomorrow with a triumphant announcement to my dev team, only to have to take it all back about fifteen minutes later.
Google Analytics Subdomain Tracking Not Working pt2
08.13.2011, 2 Comments, Google Analytics, SEO, by david gallmeier.This post is a followup to Google Analytics Subdomain Tracking Not Working pt1.
Previously, I showed you how to do a quick audit of Google Analytic’s cookie behavior, specifically, to identify a common cross site tracking issues. Today, we’ll take a look at how to fix what seems to be a common issue (here) with Google Analytic’s cross site tracking: excessive self referrals and direct traffic. The fix will involve configuring Google Analytic’s javascript in some very unintuitive ways. While this fix is not ideal, it will properly attribute referrals as well as any site goals you may set (such as checking out).
Please note that I am not a Google Analytics guru. The solution I’m sharing here worked for me and if you are grappling with the same issue, hopefully it will work for you. But no promises.
According to Google’s documentation, we want to configure Analytic’s javascript in the following way:
var _gaq = _gaq || [];
_gaq.push(['_setAccount', 'UA-12345-1']);
_gaq.push(['_setDomainName', '.example-petstore.com']);
_gaq.push(['_setAllowHash', false]);
_gaq.push(['_trackPageview']);
In particular, we’re told to insert a leading ‘.’ before our domain and _setAllowHash should be configured to false. Unfortunately, this didn’t work for me and it’s not working for you.
What did work was removing the leading period before the domain, setting ‘_setAllowHash’ to true as well as adding another line to ignore referrals from nationalbuildersupply. Here is the full Google Analytics example (make sure you include the javascript html tags):
var _gaq = _gaq || [];
_gaq.push(['_setAccount', 'UA-5971864-1']);
_gaq.push(['_setDomainName', 'nationalbuildersupply.com']);
_gaq.push(['_setAllowHash', true]);
_gaq.push(['_addIgnoredRef', 'nationalbuildersupply.com']);
_gaq.push(['_trackPageview']);
_gaq.push(['_trackPageLoadTime']);
(function() {
var ga = document.createElement('script'); ga.type = 'text/javascript'; ga.async = true; ga.src = ('https:' == document.location.protocol ? 'https://ssl' : 'http://www') + '.google-analytics.com/ga.js'; var s = document.getElementsByTagName('script')[0]; s.parentNode.insertBefore(ga, s);
})();
This must be configured on all subdomains you wish to track. For example, when tracking a visitor from www.nationalbuildersupply.com to secure.nationalbuildersupply (for checkout in this example), the javascript for both sites should be set to the above.
While this will solve the self and direct referral issues and properly record visitor traffic, we are essentially disabling cross site tracking and configuring our primary domain and subdomain to simply report themselves as ‘nationalbuildersupply.com’. Furthermore, this WILL NOT WORK for cross domain tracking – when you want to track a user from ???.nationalbuildersupply.com to ???.someotherdomain.com. Browser security does not allow a site on abc.com to modify xyz.com’s cookies (if it could, then any site you visit could lift your personal cookies for any other site, such as bankofamerica.com and facebook.com).
Additionally, tracking on your subdomain will double count visitors. Unfortunately, there is no javascript fix to this. Instead, Google Analytics filters need to be configured to ignore this (and yes, there are problems with doing this as well).
You can view a working example on nationalbuildersupply.com. Just crack open the source html and search for: var _gaq = _gaq || [];. With these settings, we were able to track visitors from www.nationalbuildersupply.com through the checkout process, which takes place on secure.nationalbuildersupply.com.
I’ll follow up with more material explaining the inner workings of Google Analytics in further detail, including implications and other limitations. Included will be filter workarounds.
Google Analytics Subdomain Tracking Not Working pt1
08.13.2011, No Comments, Google Analytics, SEO, by david gallmeier.You’re tracking users with Google Analytics. While GA is simple enough to deploy on single domain sites, watching visitors’ traffic across domains and subdomains gets tricky fast. I’ve seen a number of posts from users experiencing excess amounts of self referrals, direct traffic and the like. In theory, Google Analytics enables us to filter traffic like this. In practice, getting GA to behave properly can be a real headache.
Case in point: you’ve followed Google’s documentation on the various tracking scenarios to the T and it’s… Just. Not. Working.
I’ve run into this problem recently and hopefully this series will show users how to test that GA is in fact functioning properly by monitoring cookie content (it’s not that hard) and show them how to address issues that come from cross-domain tracking. Also, I may not have the optimal solution to this issue. In that case, feel free to chime in with a better fix.
So today, we’ll start with a simple audit that anyone with Firefox can perform, to ensure Google Analytics is functioning properly.
The first step is to install the Cookies Manager+ addon for Firefox. This step isn’t completely necessary, since every browser will display cookies, but I find checking GA’s cookie behavior easiest with plugin like this.
Once done, head to your site and open Cookies Manager+. Lots of cookies will probably show up, the majority of which don’t matter to us right now. Filter these out by typing your domain into the search bar at the top.
These are your cookies and if you are having problems with Cross Domain tracking, the cookies have probably gone bad. I’m assuming you’ve got organic ranking at some point, so delete all the _utmXX cookies associated with your domain (these are the Google Analytics cookies) and click in from an organic or PPC result. You should now have a freshly baked batch of cookies, which looks something like this:
In particular, we want to watch _utmz. This cookie stores the referral information. Whether you’ve landed from organic search, PPC, direct traffic (bookmarks, email campaign or typing in the URL), site referral, etc, _utmz will record the source of your traffic.
Here are some _utmz examples of direct, self referral, organic and PPC cookies, respectively:
Direct traffic
84761990.1313282840.1.1.utmcsr=(direct)|utmccn=(direct)|utmcmd=(none)
Self referral
168166435.1313282936.1.1.utmcsr=nationalbuildersupply.com|utmccn=(referral)|utmcmd=referral|utmcct=/Cart.aspx
Organic
84761990.1313283166.2.2.utmcsr=google|utmccn=(organic)|utmcmd=organic|utmctr=toto%20toilets
PPC
84761990.1313283206.4.4.utmgclid=CPTtqNXIzaoCFUjt7QodTnDczg|utmccn=(not%20set)|utmcmd=(not%20set)|utmctr=national%20builder%20supply
The prefixed numbers aren’t important for what we’re analyzing. They record ID’s, visit counts and related metrics.
What you do now depends on the layout of your site, but you essentially want to browse the majority of pages on your site, monitoring the _utmz cookie content. You are looking for any change from organic/PPC/other site referral to self-referrals and/or direct traffic. Simple browse through your site and watch the content of the _utmz cookie.
If your cookies are suddenly dropping important information, like organic sources or PPC referrals, to report a self referral or direct traffic, you’ve got a problem. Basically, Google Analytics is overwriting important referral information with direct and self referrals (something Google Analytics should never do). This seems to be a problem when trying to track visitors across multiple domains and subdomains.
To view the fix (well, what fixed GA in my case), see my next post: Google Analytics Subdomain Tracking Not Working pt2



