Advanced Segments in Google Analytics – Examples to Get Started

Advanced Segments in Google Analytics solve a problem for me!

For all of my forward thinking in designing and tagging a client’s website to track and report on key performance indicators, inevitably there will come a need to dissect historical data prior to custom variable or profiles being implemented.

Advanced Segments Pulldown

That’s where the beauty of advanced segments (released by Google Analytics in late 2008) comes in to save the day.  This feature allows users to pull out subsets of their visitor data and view them as though they were the only data within all reports of Google Analytics.

For example, you already know from your default dashboard (shame on you for not customizing it yet for your needs, not those of the masses) what percentage of your visitors comes in directly (or through search or through referrals).  But wouldn’t it be nice to have many of the reporting tools available to you for just that advanced segment?

Google Analytics has useful predefined segments:

  • All Visits (the default  – what you’ve been viewing since Day One)
  • New Visitors
  • Returning Visitors
  • Paid Search Visitors
  • Non-paid Search Visitors
  • Search Traffic
  • Direct Traffic
  • Referral Traffic
  • Visits with Conversions
  • Mobile Traffic
  • Non-bounce traffic

These are some great advanced segments to get you started.  But the real value comes when you define your own segments tailored and personalized in a way no package could predict.

Examples from clients

Some were used one time to track down an issue or answer a single question; others are used regularly as part of my key performance indicator (KPI) strategy:

Here’s my shorthand: (Advanced Segment Friendly Name) = (syntax of the advanced segment definition)

  • Google Analytics Segments ListBlog Pages = Page starts with /blog
  • Indian Visitors = Country/Territory Matches exactly India
  • Non-Indian Visitors = Country/Territory Does not match exactly India
  • ABCkeyword = Keyword contains ABC
  • Keyword is DEF = Keyword Matches exactly DEF
  • Keyword is not DEF = Keyword Does contain DEF
  • Google Searchers = Source Matches exactly google
  • GHI LP = Landing Page Matches exactly /GHI
  • JKL conversions = Goal4 Completions Equal to 1
  • Campaign MNO = Campaign Matches exactly MNO
  • 404 Page Not Found Advanced Segment404 Reports = Page Contains /404.html
  • Visits With Site Search = Site Search Status  Matches exactly Visits With Site Search
  • Non-Customers = Page Does not match exactly /AAA/BBB
  • Non-Customers Non-Paid = Page Does not match exactly /AAA/BBB AND Medium Does not match exactly cpc
  • Firefox Browsers = Browser Matches exactly Firefox
  • Goal Completions = Goal1 Completions Greater than 0 AND Goal2 Completions…
  • Goal Starts = Goal1 Starts Greater than 0 …

BOLD ITALICS LETTERS are client-specific terms.

Each of these advanced segments can be turned on and combined with others to compare and contrast.  The power of this level of reporting is huge and should be exploited.

Once you’ve mastered Advanced Segments you’ll move to Custom Reports, Intelligence and then really get into the tagging of your visitors, events and conditions during their visits.  This will allow you to perform deep forensics and pull data and information that will drive fundamental business decisions.

Twitter Referrals and Web Analytics – A Broken Referral Link

Broken Twitter Referral Link

If you are obsessed about your web analytics or your customer’s web analytics as I am, then you may have noticed a problem where Twitter referral traffic is being recorded as direct entry traffic rather than referral traffic.

Don’t mess with my numbers, man!

I work hard to keep my numbers clean.  I do it for my sites, my employer’s sites and my customer’s sites.  Without this anal retentive attitude you cannot make higher level business decisions.  The supporting data is flawed so your assumptions are broken.

But that’s a rant for another day.  Just suffice it say that you need to constantly test your data to make sure it’s legit.


Who Can I Blame?

Web Analytics Referrals From Twitter

Here’s the problem in a nutshell. When you click from one web page to another the browser usually passes referral data to the receiving page.  That data is then recorded by your web analytics program so you can report on where your visitors arrive from.

In Google Analytics they break it down into 3 buckets initially, Direct, Search and Referral.  Now, if I tweet this blog post’s URL through Twitter, I want those click-throughs back to the site to be recorded as referrals from Twitter. Likewise when others retweet me I want them to also be recorded as referrals, not direct entries.

But the Twitter model has introduced a new presentation screen different from browsers.  HootSuite and TweetDeck are popular applications used to “dashboard” Twitter activity (along with Facebook and LinkedIn).  These applications and their tight relationship to URL shorteners, do not always pass the referral data (needed by web analytics tools) you would normally see if they had come from the twitter.com domain via a browser.

If you want the down and dirty details behind it visit Danny Sullivan’s forensic work on it.


Make The Numbers Match!

Connecting Twitter Visitors In Web Analytics

Fine.  It’s a problem.  But you need to solve it, right?

Again in a nutshell: force the referral information to be preserved as visitors click through Twitter to your website.  This is done with link tagging.

  • Using Google Analytics, we can go to their URL builder and force-tag our link before we put it into our tweet.
    • Of course that gets really painful if you tweet more than once a week.  So check out Snip-N-Tag for an inline method of adding link tagging for Google Analytics.  Pain relieved!
  • Using Omniture’s SiteCatalyst, you can create a campaign in your report suite for all of your Twitter postings.  Then append the campaign id (e.g. s_cid) to every one of your tweets.  You’ll also need to further manipulate some of your variables to ensure they’re attributed to referral traffic, but that’s beyond this post.

I always try to include link tagging on every link I place out there.  Even ones that are not destined for one of my sites.  Nothing speaks to an analytics guy or gal more than looking in their report and seeing your traffic to their site jumping off the screen with campaigns names.

I should really share that treat here!  Next week.

Web Analytics for Banks, Three Reasons to Start Today

Web Analytics For Banks

This one I don’t get!  There aren’t many, but I still run across a few banking websites not running web analytics.  Credit unions are the worst offenders, by the way.

Why is this?

Why not throw Google Analytics on there for free?  If nothing else, have it there knowing you’ll always have a benchmark should you ever need it one day. Which leads me right into…

The three reasons to implement web analytics today!


Benchmarking Your Web AnalyticsWho’s Been Sleeping In My Bed?

1. Benchmarking

I’ve said before in posts, if you don’t have web analytics installed on your website then don’t talk to me.  When I am asked to come in for an assessment I need data.  Sure I could pore over log files, but that takes a significant amount of time (the banks’ money) and rarely does it allow me to push up the data in a manner needed by my customers to make the higher level business decisions.

One day you or a consultant is going to be called in to answer some very tough questions about a past event or what to expect during an upcoming seasonal event important to your financial institution.

No web analytics = No insight = GUESSWORK


Attribution Of Your Web ChannelsWhose Phone Numbers Is This?

2. Attribution

  • Do you know the makeup of search traffic vs. referrals vs. direct visits?
  • Are your targeted SEO keywords actually the ones driving the most traffic?
  • After your homepage, which is the most common first page?  Does that page represent your banks in the best light?

This list goes on.  You get the point.  Even the simplest of analytics implementations allows you to go back and answer questions you didn’t know you needed to answer.

One of my favorite questions I pose to my clients.  “Which site drives the most traffic to you?  No wait, which site drives the most converting traffic to you?  Oh, it’s the ACME School District in the neighboring county?”  Well it’s time to see what’s going on with their website and thank the author of that inbound link, but not for the SEO value of the link.  I’ll take a converting link over a hundred non-converting links all day.

But how would you know?


Action-Oriented TacticsLet’s Be More Than Friends!

3. Action Time

You’re ready to move to that next step with your website, right?  But what does that mean?  Can banks really actively impact online banking customers?

It should mean that you’re ready to confidently move those website KPIs you defined for your site and your marketing plan for the bank.  Now is when the analytics for banking really kicks in.  You will start tagging your campaigns, building those advanced segmentation reports and tweaking those budget allocations.  From there you will grow into the next level of testing, redesigning and experimenting with new concepts.

The nice part of the “action time” is the immediate feedback you’ll receive.  You can quickly can the losers and mutate the winners into more powerful tactics.

This iterative process will put your bank years ahead of your competition in a matter of months.


Addictive Web AnalyticsThis Is Some Serious Crack!

Web analytics gets addictive.

Well maybe to those of us who live in that lonely chasm between marketing and technology/IT.  I love it here.  The impact a smart business analyst can have on your bank’s bottom line is an order of magnitude more than you could ever spend on the software needed to empower that analyst.

But don’t wait for that analyst to be hired or contracted.

Install the analytics now so when he does come on board he can quickly knock it out of the park.

New Google Analytics Tag – Faster, Better, Stronger

Google AnalyticsYesterday, Google announced their new Google Analytics tag to speed up page loads that will become the default code snippet provided for profiles.  This is right in line with Google’s new obsession with page speed.

Google Analytics New Asynchronous Code Snippet

While the new Google Analytics tag is very light, if you start to add a lot of scripts to a site like I do, they start to add up.  Allowing for the asynchronous loading (separate processing) of the tag your webpages will load that much faster and therefore the code snippet will not penalize your site speed nor your visitor’s experience.

I recommend you go back to all of your Google Analytics-tagged pages and change them out to this new snippet.  Likewise, you really need to look into the load times of your webpages.  It could be hurting your rankings.