skip to main content

TAMU Webmaster's Blog

Information and insight from the A&M Webmasters


Data Driven Decisions Made Real

Last month GoWeb sponsored a presentation on campus for Intermediate to Advanced Analytics.  The director of the university’s visitor center (which is part of our division) was interested in learning more about how to look at and analyze the site’s traffic, so attended the session.  After getting back to the office the next day she came back even more interested in being able to do a deep dive into the site.  What we found demonstrates exactly why analytics are so important, and why making decisions based on the numbers is crucial.

At the presentation we talked about the importance of establishing a goal and setting up conversions within Google Analytics (GA.)  This was relatively straightforward for the visitor center – their primary goal is to have someone register for a campus visit.  So we set that up as a conversion and looked at the traffic funnel that drove people to that registration.

Coincidentally we had just redesigned the entire Visitors section of the university website a few weeks before.  It had been carefully done, the content rewritten, design thought out, and vetted by everyone up to the vice president level.

The numbers were shocking.  Registrations to individual visits were down by 35%, and registrations to group visits were down by almost 80%.  The date of the change was obvious by looking at the sessions per day graph.  Also coincidentally, our vice president had recently discussed the idea of assigning a monetary value to a campus visitor, which makes the change in site traffic hit home even more.

In retrospect, the causes for this decline were obvious.  Rather than the new design making it easier for the user to accomplish their goal, it actually created several new barriers that did not exist before.  It was a great example of including what the project sponsor wanted it to include, without taking into account what the end user wants to experience.

After talking about these issues we quickly pulled in the visitor center team, web team, and design team to discuss a solution. As a quick fix, the call to action buttons were places more prominently on the page. A followup project will bring the design and development team back together to create a more streamlined solution.


Friday, October 20th, 2017 Analytics No Comments

Google Analytics Documentation Available

We tried to use last week’s GoWeb presentation on web analytics as a way of re-energizing our campus-wide discussion on analytics.  As part of that, the GoWeb analytics special interest group (SIG) has updated our documentation site.

These pages will explain what Google Analytics is all about, what we are trying to accomplish with the campus-wide effort, and give instructions on how to set up and manage your own analytics. Whole books have been written on this topic, so we are obviously presenting an overview, but we hope that this will be a good start for those of you on campus who want to become more active in your own analytics use.

Wednesday, October 4th, 2017 Analytics No Comments

Blocking Triggers Refined

In my last post about blocking triggers I showed how you can use exceptions to prevent Google Tag Manager from firing particular tags.  Because the HotJar tag contains the service ID value for that site we had to create individual tags/triggers for each site.

But what about when you have a more generic tag that is not linked to a particular service ID?  For example, you want to include an HTML or javascript snippet, but you still want to be able to control which sites it is or is-not displayed on.  You could create individual tags/triggers for each site as we did with HotJar, but there is actually a better solution.

The process starts the same, by adding a variable to your GTM.  This time, though, we will use the “Custom Javascript” variable type instead of the “Constant.”  The javascript we use then sets the URL of all of our sites that we want to exclude into an array.  It then loops through the array and tests whether the page being viewed is in that array.  If it is, then the function returns “true.”

At that point we can continue the same process that we showed for the HotJar tags.  Add a trigger that executes when the variable value returns “true” and then make that trigger an exception.  Now you are excluding that tag from firing only on the sites that you put in your javascript array.  To add or remove sites, just update the array in the variable and republish your tag workspace.

There are several other variable types. I have barely scratched the surface.  DOM Element looks particularly interesting.  This or another combination might actually solve the problem of needing duplicate tags for our HotJar implementation.  I will let you know if we ever get that far.


Tags: , ,

Wednesday, August 23rd, 2017 Analytics No Comments

Blocking Triggers in Google Tag Manager

In looking at how to optimize our sites download time, one common recommendation is to combine all javascript into a single file.  An even better piece of advice – if there is javascript that your site doesn’t need, don’t download it in the first place.

One of the areas that we identified as being bloated was how we ran our HotJar heat map implementation through Tag Manager.  HotJar tags require an individual id number corresponding to the site being monitored.  Since we have several dozen sites, that means several dozen tags.

The default for tags is to run on all pages.  This meant that all of our sites were firing off all of the HotJar tags, regardless of whether they were even for the proper site.  Realistically that might not amount to much, but I knew we could do better.

One solution might have been to create different triggers for each site, but that would have quickly become unwieldy.  After stumbling across an article talking about trigger exceptions I decided to go down that route.  I first created a variable “Block Sites” with the value of “block” (it could have been “1” or “true” or whatever.)  From that I created the trigger “Block All Sites” that consisted of the simple comparison “When Block Sites = block” (in essence, “when 1=1”.)   The key is to then add this as an exception to your exiting tag firing triggers.  So the tag fires normally if the exception is not present, but does not fire at all if the exception is present.

In theory, since exception prevents the tag from firing, your page never downloads and runs the code associated with the tag.    We do not continually run HotJar, so we can keep the exception in place until we are ready to start collecting data. Then all we have to do is edit the tag to remove the trigger exception and publish the new version.

I do believe there is another way that eliminates even this through the use of an all-javascript trigger, but I’m not myself adept enough with javascript to get all the pieces put together.  I will post at least the first part of the process in a later post that doesn’t include the requirement to match the HotJar id value.

Tags: , ,

Thursday, August 17th, 2017 Analytics No Comments

Bowl Game Day

Throughout the course of the year I had been watching the university site’s traffic on gameday.  I have mentioned how gameday does affect traffic – not only the amount but also the pages people are looking for.  I therefore decided to use this trend to try to encourage visits to some of our other web properties. On bowl game day we therefore swapped out the “In the Spotlight” section on the front page of the university site with links to Aggie Traditions, 12th Man, and Reveille.

My theory was that these sites all receive more hits during game weekends without any additional advertisement, so adding them on the front page should drive even more traffic.  The numbers, though, are inconclusive at best.  Each of these sites did – as expected – experience an increase in traffic.  The increase was not as much as I would have thought, though.  More telling, not as much of the increase came from referrals than I had hoped.

The university site did experience an increase in traffic, but not as much as it did for some of the bigger games during the season.  This would have meant that we didn’t get as many people seeing the links to have clicked on them.  Organic search still dominated as the channel by which people were landing on these sites.

We also didn’t attract as many outside visitors to the university site as with other gameday peaks.  On average we had 42% new sessions during the semester.  The October 8 game against Tennessee, fueled by our also hosting ESPN’s GameDay, saw a peak of 78% new sessions.  For the bowl game it was only 62%.  This percentage drop, coupled with the lower overall number of visitors, would imply that there were fewer visitors who might have been looking for this type of information.

The game itself may have lent itself to these trends.  Having been in the Big 12 for several years, most of the Kansas State fans are probably familiar with Texas A&M.  The matchup probably didn’t have the national draw that some of our in-season games did, so we would have fewer people from areas of the country who aren’t as familiar with who we are.

So while I don’t think this was a failure, it still wasn’t the success that I was hoping for.  It does indicate that we need to be aware of scheduling these content changes around events that can get more outside attention than we were able to draw this year.

Thursday, January 5th, 2017 Analytics No Comments

Google Data Studio

Google recently released a limited access to Data Studio, which allows us to create highly visual analytics reports.  Their product description page shows two examples of how different – and how impressive – the reports can look.  We all know that presenting the standard Google Analytics dashboard to our decision makers is largely ineffective.  There is simply too much information formatted in a hard to read manner.  Data Studio can make your analytics effective simply by making them read by the people who need to be reading them.

Creating a Data Studio report is similar to using such services as Gliffy or Visio.  You can select whatever metrics that you want to display, position them on the page, and style them to reflect your own preferences or office branding.  Almost any metric that you can access through the GA dashboard should be accessible from within the Data Studio interface.

I have recently created a few reports and given them to my decision makers, and the results were beyond spectacular.   One administrator spent several hours looking at the Data Studio report on the day that I sent it, probably more time than had been spent in total on the typical dashboard reports that I had been sending for two years.  This review resulted in a meeting of that particular project team and decisions being made to change plans on how to proceed on it.   The idea of “data driven decisions” suddenly made sense…because now they had data that was accessible to them.  Of course this result will not be the typical response, but it does show that repackaging our analytics data into easy to read reports is an effective way of increasing its perceived value and even the amount of attention and use it receives.

Friday, July 8th, 2016 Analytics No Comments

Adding “Email Opened” to Google Analytics

One of the basic claims of many bulk email platforms is the ability to track when the email is opened, which they call the number of “opens.” They usually do this by adding tracking codes or transparent images that are tied to the recipient’s email address. The Maestro system on campus does not have this feature activated, which is probably a good thing. At the same time, we have not been able to gauge how many of our readers open the Texas A&M Today newsletter when they receive it.

Even those platforms that do keep record of “opens” do not always make this available to your Google Analytics.  Not a requirement, but it is convenient to have all of the information in one place when you start to look at your site usage.

By adding campaign tracking codes to links in the email, we were able to see that many of our news stories became popular just because they were featured in the email newsletter. But until now, we couldn’t tell how many people open this email newsletter when they receive it.

We recently built a method within Google Tag Manager that does allow us to track when (but not by whom) our emails are opened… at least if we define “open” as opening the mail and downloading images. Using the Google Analytics Measurement Protocol, we created an “image” (really just a URL) that stores an Event Category of “email,” an Event Action of “open,” and the title of the news story as the Event Label. This works because the email thinks it’s a real image and tries to download it, sending the information to Google Analytics.

Our analytics show that recent issues of the Texas A&M Today newsletter were opened by more than 20,000 people the first day they were sent, 3,000 the next day, and 1,000 the third day.


Tuesday, July 5th, 2016 Analytics No Comments

Random Findings in Analytics

Ask a random student on campus where they are from and it has always seemed like the majority will say they are from Houston.  Coming from the Dallas area myself, and knowing how many of my own high school classmates came to A&M, that has always seemed surprising to me.

Looking through our analytics, though, the numbers support that anecdotal conclusion.  One of the pieces of information that you can get from analytics is the geographic location of the user – from the country level all the way down to the city.  While there is some margin of error when you look at the by-city numbers, they are accurate enough to make some comparisons.

Unsurprisingly, College Station is at the top of the list, with just over one-third of traffic coming from here.  That makes sense given the amount of student, faculty, and staff traffic that we get.  What did surprise me, though, was how much more more traffic we get Houston than anywhere else in the state.  Just over 15% of traffic comes from Houston – more than Dallas, Austin, and San Antonio combined.

These numbers are fairly consistent across all of our sites.  I have started watching the combined view of all university sites that have joined in on our campus-wide tracking code and there isn’t any change after adding several colleges, departments, and offices.

Whether it is proximity, the number of old-Ags living in Houston, or some other factor, the percentage is significant enough that it should be taken into account if we ever do location-based advertisement or programs.

Monday, June 6th, 2016 Analytics No Comments

Adding campaign tracking codes on the fly

Campaign tracking codes don’t require advance setup or technical expertise. Any writer and marketer can and should include tracking codes in the links they send or tweet out, especially if it’s an ad they’re paying for. That way, using Google Analytics, they can see which of their efforts is most effective.

Here’s an example of a link:
(Okay, that part was easy).

Here’s an example of a link with tracking code:

The code may look complicated, but here’s the breakdown:
=twitter or facebook or tribune and so on.
utm_medium=social or email or banner or adwords and so on.
=12thman-Launch or 04/08/2015-email and so on.
=tweet-graphic, website-link or get as specific as you want, to see which link or ad brought the most traffic.

Even better, use & instead of &. It’s more correct and may avoid problems.

To analyze your campaign results, log into Google Analytics and, from the left menu, choose Acquisition > Campaigns > All Campaigns. Then  under Primary Dimension, you can look at your data in different ways:

  • Campaign
  • Source
  • Medium
  • Source / Medium
  • Other


Friday, June 26th, 2015 Analytics No Comments

Webmaster Tools and Their Limitations

What tools can webmasters use to understand, as Monty once asked, who our “web visitors are and what they want to see”?

The Google Keywords Planner (under Tools in your Google Adwords account) can tell you what related keywords are most popular worldwide. Google Trends looks promising – compare search popularity by time and region for colleges or english majors. Google Analytics can tell you where your visitors came from, the keywords they were searching for (if Google feels like telling you), and what pages on your site are most popular. For search, Google Webmasters Tools may tell you even more. Your server logs can tell you some of those same things too.

But none of these tools can tell you about the people who have never visited your website.

So, what’s a webmaster to do? Blindly rewrite all our pages to use all the most popular keywords? Google Trends has a list of those too. Blast an email to every website that’s vaguely related to yours, and try to convince the webmaster that your site is closely related to theirs and needs a prominent link from their site?

I’m a little sensitive about link building requests, after recently receiving two emails from the same “link building marketer” on behalf of two normally-reputable educational publishers. He was asking for his links to be added to some of our web pages.  First to the Key Public Entry Points page, except that contains only A&M home pages. Then to one of our news releases, except that was released to the news outlets three years earlier. Apparently he didn’t actually read the pages first.

No, don’t turn your site into something that it isn’t, or pretend it’s something that it isn’t. That doesn’t mean that you shouldn’t let these tools serve as a wake-up call. If nobody is linking to your site, you probably need to start answering the questions that people are asking. If nobody is interested in what you’re talking about, eloquent and informative as you may be, you probably need to start talking about something new.

When our customers change, our products need to change too. When the world changes, departments should change. But a departmental website won’t succeed unless it accurately reflects its department. The department has to change first, then the website.

You’ll always be most successful by being yourself. For one thing, there’s no competition. Once you’ve decided who you want to serve – and who you can serve best – then give those people what they want to see. How do you know what they want to see? You have to enter their world. Or you could ask them.

Tags: , ,

Monday, June 16th, 2014 Analytics No Comments