Embedded content is lovely! Here’s some now…
Before you consider using the data in your Google Analytics account to start seeking insights and drawing conclusions, there are a number of things you should check to see if they’re being handled correctly/optimally.
This is not an exhaustive list, but in preparation for a webinar I’m doing on Google Analytics, I started putting together a number of common pitfalls – configuration and tracking details that could potentially introduce flaws in your Google Analytics data.
is everything you have on a single domain?
is it OK to count traffic from your staff?
do your URLs have parameters that make pages that are the same seem unique?
do you know what those URLs look like?
Checking that you’re tracking:
Use the Chrome browser, and add the plug in called ‘Google Analytics Debugger’
Navigate to your site
Hit F12 to open the Console
Activate the GA debugger plug in
Navigate the site and see what data is being sent to Google in the console. This is pre-filtered data.
When trying to set up Google Analytics Conversion Funnels, it’s helpful to know what URLs make up the sequence of pages someone goes through to accomplish a task.
For TeamRaiser by Blackbaud (nee Convio), the URLs in the funnel for making a donation to a participant are (ignoring domain) as follows below, using common regular expression patterns. There are four of them and they only differ by one parameter.
You can simplify that URL by having your profile ignore certain CGI parameters, like PROXY_ID and PROXY_TYPE, as well as JServSessionIdr004, idb, fr_id and df_id, but give that some thought before you do so.
The second, third and fourth URLs look very much the same, except for the donation parameter:
In preparation for a presentation on how to increase ‘engagement’ on a website, I assessed a number of health-related non-profit websites for how they made use of a variety of features.
Here are the features we considered:
And the sites:
Global Fund for Women – responsive design, lots of photos, video sharing campaigns, hero sliders, Pinterest and other soc med, rolling asks on thank-you pages
San Francisco AIDS Foundation – Facebook commenting, hero sliders, photos, addthis sharing, content cross-linking
Health Initiative for Men – Super gold star for interactivity! Fun ‘quiz’ to help people identify their risk level
Greater Than AIDS – heavy use of imagery, clearly representing their target ethnicity; huge homepage slider; floating social sharing widget
Food Water Watch – Factory Farm Map – Really rich, interactive data-driven map that presents as very simple, easy to use.
World Wildlife Fund – great responsive design, gorgeous photography, all content shareable, How You Can Help feature on every page
Charity:Water – terrific use of infographics and video (Why water? Page)
Malaria No More – modal window to promote current campaigns, elegant simplicity, tells the user exactly what they should do
Below are some of the the websites and blogs that I have found to be particularly useful in building an understanding of Key Performance Indicators (KPIs) and how to apply them to web analytics, particularly for non-profits.
Avinash Kaushik writes extensively about web analytics. He has three books out on the subject, and maintains a blog with very rich, informative posts that always emphasize the concept of ‘actionable insights’.
Particular posts pertaining to metrics:
KISS Metrics offers a product for measuring web traffic, but they also offer a free and often interesting blog on a variety of related topics. This page has articles specifically on analytics/metrics:
Shabbir Safdar is a Bay Area consultant who specializes in working with non-profits, and who has a great reputation. He blogs about a variety of issues pertinent to non-profit web work:
Jeff Shuck, of Event360 is incredibly passionate about data analysis. If you ever get the chance to see him in person, do it. In the meantime, he sometimes blog about data:
M+R Strategic Services is also a consulting vendor. They’re noteworthy because they produce a benchmark report every year reflecting a variety of metrics across the non-profit spectrum:
When you want to know where your website traffic is coming from, as far as links that you are promoting from emails, web ads, paper mail, etc, Google Analytics provides a lovely tool that makes it easy to create URLs that include that particular source info.
The results are that instead of seeing 10 variations on ‘xxx.yahoo.net’ in your Referrer reports, you can see that the traffic actually came from the third text link in the second paragraph of the email that you sent to your advocacy sub-group on March 3. I’m disregarding why on earth you would have three links in a single paragraph when you’re emailing your advocacy sub-group…
But this post is about how to use the tool better. Continue reading
Having finally switched to gmail as my main email client, I searched for years for a way to load my years-worth of older emails into gmail, while maintaining dates and other meta-info (like sender).
The blogger at swampgas.com made a python script that does exactly what I was looking for, and it will actually work to transfer any mailbox using the traditional mbox format into an IMAP server.
Awesome-town! It takes a while to run, especially for large files, and there’s a bunch of log output (which is good). But I have a lot of old files, so I didn’t import them all at once. i also had to run the script multiple times on some files, because it would crap out on a particular message.
I found that I would forget exactly how to run the script, and where to put the files, etc etc etc. So I tweaked his code a bit, and decided to add some more command-line options as well.
download the Python code and sample config file here.
NOTE that this is not really for non-coders. You have to be able to run a Python (version 2.5 or better) script that has a number of modules installed and (obviously) internet access. There are some simple parameters in the config file, but there are a few things that are hard-coded as well that you might need/want to adjust as I did.
I’m not a pro Python coder, so it’s doubtful that I’ll be able to help debug problems.