April 27, 2015 / by Jim Dougherty

I started writing this piece with the intention of sharing “best practices” for using Google Analytics to measure PR. It might of been a pretty helpful post for some people, but it could have been very counterproductive for some as well.

If you’re using a physical tool (a hammer for example), you would probably decide to use it when you had a nail to drive. It would be unusual to decide to buy a hammer and then see if you could find something to pound with it.

But that’s what we do with a lot of PR metrics: we measure impressions, social sharing and AVE because they are easy to conjure, not because it has applicability to our overarching business objectives. Most PR metrics are hammers without anything to pound. Google Analytics without an understanding of its mechanisms is useless hammer.

Shonali Burke, one of the eminent thought-leaders in PR measurement calls these tools “shiny objects.” We’re drawn to them, but they don’t have a lot of applicability to what we do.

I’m going to assume you’ve done the hard work of understanding how your PR activity is supposed to drive your business objectives. What I want to do in this post is to break down the functions of Google Analytics to help you decide if the tools can help you measure what you want to measure.

What’s the difference between free and premium?

The first question you might have before we dive into Google Analytics is “how much of GA is free?” That’s a pretty fair question – so let’s explore the differences between premium and free Google Analytics accounts.

First of all, GA premium prohibitively prices smaller users out of the enterprise solution. The annual licensing fee for GA Premium is $150,000.
GA premium also may be available at a discounted rate from authorized resellers, but it’s clear what size of enterprise that this solution targets.

Basically, the free version of Google Analytics is a self-serve product that measures up to 10 million “hits” per month and is updated every 4 hours.

Premium offers 24 hour support, measures up to one billion “hits” per month, has dedicated account management, data update < 4 hours, product training, and integration with the DoubleClick product ad exchange.

Sufficed to say, this post is focused on the free version. 🙂

How Google Analytics collects data

When we discuss Google Analytics you may think about reports you’ve seen or reports other people say you should see. And while reporting is the sexiest thing about GA, it’s important to understand where this information comes from and how it’s processed so you can vet a report’s utility.

It’s a pretty well known fact that most statistics are calculated correctly but draw calculations from weak (non-representative) data sources Understanding how data is collected and processed through GA will help you to have confidence that you’re getting good data to inform your decisions.

When you configure Google Analytics for your website, you add a javascript code to the head of your HTML document. The code triggers an asynchronous data collection sequence when a browser renders HTML from your page.

The data that the javascript sequence collects is organized by Google into a three-part hierarchy:

User – this is a unique identifier assigned for a browser for each device. This information is retained as a first-party cookie in the browser unless a user clears their cookies, at which point a new “unique” identifier is generated. You can modify the unique identifier feature to use a custom identifier (such as a unique user ID for your site) that purports to track unique users across multiple devices as well.

Session – these are the individual visits to a site. A visit starts when the script starts loading (when a user first visits your site) and continues until a user has been idle for 30 minutes (this is called the “timeout length” parameter). Google says this is a sound parameter for most purposes, but for a platform like Yahoo Screen or any site showing videos for 30 minutes or more this parameter would probably need to be modified.

Interactions – these are the activities that happen on your site which Google calls, “hits.” Everything from the pages visited to the links that they click are captured and associated to sessions and users.

Note that there are two javascript libraries for Google Analytics: classic Google Analytics (ga.js) and Universal Analytics (universal.js). For the purposes of this post we’ll be discussing Universal Analytics, which is the more robust (and preferred) option of the two. (If you need to upgrade from classic Google Analytics you can read more about how to do this here)

For mobile apps, this process is replicated more or less by a software development kit (SDK) rather than a javascript library. The differences between how app data is handled compared to javascript are that app data is batch processed (to ensure connectivity and limit use of mobile battery power) and the app data unique ID is regenerated only when an app is deleted and reinstalled.

There’s more about the iOS and Android SDKs here, but I’m going to concentrate on GA for web properties for the remainder of the post.

You can also manually import different types of information into Google Analytics. There are two main types of manual import that Google identifies:

  • Account linking: this is the linking of other Google properties (AdWords, AdSense, Webmaster Tools) to your analytics data.
  • Data import: This would be the import of data via spreadsheet of .csv file. Examples of this would be “dimension widening” where you add context (such as targeting data, or content authors) associated to content on your site, and cost data import which would add performance data to GA for non-Google ads. The literal key to this is a database key that associates your new data uniquely to “hit” data.

Another really great tool for Google Analytics input is the Google Tag Manager. Google Tag Manager allows you to manage tags on your web properties that can give you much more detail than GA (and it can inform remarketing and other tools requiring specific behavior tracking mechanisms). I’m going to disregard it for the purposes of this post, but am including a video so you can get a sense of some of
the granular tracking back to GA that you can accomplish with this tool.

How Google Analytics processes data

When the Google Analytics javascript generates an image request to the Analytics servers (this is the “hit”), it returns a file with a bunch of condensed, raw information. Google Analytics has to parse, process and transform that data into database fields that are useful for report generation. Understanding this process is important to making sure that your data is the data that you want.

Your settings have a lot of power to increase the quality of the data that you use in GA. There are three key actions that you can perform
in to shape the data that is returned to you in your reports:

Implement filters:

  • Include your data
  • Exclude your data
  • Modify your data

For example, if your business is only in a few key markets, then it might make sense to restrict data to those markets and filter out your Slovenian hits. GA has a complement of preset filters and has the capability to implement customized filters as well.

Set goals:

Google points out that the payoff for setting up this feature is immense. You can then track:

  • Conversion rate
  • Conversion path
  • Compare campaigns
  • Much more

By configuring these correctly you can do some really useful analysis, such as evaluating the customer path for a specific funnel (these would be “destination” goals).

Google Analytics - PR Measurement

Set Groups:

There are three ways that you can group data in Google Analytics:

It probably worth noting that data grouping isn’t retroactive in Google Analytics, which means that proper establishment of these upfront will save you a lot of time on the back end of your analyses.

How Google Analytics reports data

Google Analytics reports are probably the product that you are most familiar with and have the most experience with. They are the end-product for Google Analytics and this is what people generally focus on in their best practice posts. Hopefully because of your familiarity with how the data is collected and processed you can find ways to make your reports more relevant as well.

There are two aspects of Google Analytics reports:

  • Dimensions – these are the characteristics of the collected data
  • Metrics – these are quantitative measurements of the collected data

For a typical Google Analytics (default) report you will have table with one dimension in the left column and multiple metrics to the right. You
can always add secondary dimensions and manipulate the metrics…particularly if you’ve been thoughtful about how you set up data collection and processing for your account.

Not all dimensions can be used with all metrics, however. One thing to keep in mind is the three-part GA hierarchy that Google set-up:

  • user-level dimensions can only be used with user-level metrics
  • session-level dimensions can only be used with session-level metrics
  • hit-level dimensions can only be used with hit-level metrics

If you think about it, this makes some intuitive sense. After all, the unique visitor count for any individual hit should be one. 🙂 Here’s a dimension and metric reference guide to help vet what parameters you can measure.

For many of the default reports, the processing step makes all of data accessible and easily queried. For custom reports this is not always the case… which is why you need to be aware of sampling. In order to calculate complicated queries faster, Google Analytics will sometimes use a representative sample of the data to compute a report.

When this is done you always have a yellow prompt that shares that the report is based upon a sample of data and the percentage of data sampled. From that prompt you can change the sampling setting depending upon your preferences for speed versus accuracy.

For reporting, you also have the option of querying via one of four Google Analytics APIs to populate a dashboard or other offsite tool. This is developer-level stuff so I’m going to skip over it, but there is more information on the available APIs here.


Hopefully this is a helpful way to look at Google Analytics. You can see that it is a very robust tool that can measure a lot of things (that aren’t even on your website), but it needs to be understood and set-up properly.

I’ll close with a quote from one of the principle engineers of Google Analytics, which is very congruent with the prevailing thought about PR and marketing measurement:

“The platform is the underlying tool that makes all of your analysis possible.” – Sagnik Nandy, principle engineer of Google Analytics

Tags : measurement

Most Recent Posts

Cision Blogs Topics

  • Communications Best Practices

    Get the latest updates on PR, communications and marketing best practices.

  • Cision Product News

    Keep up with everything Cision. Check here for the most current product news.

  • Executive Insights

    Thought leadership and communications strategy for the C-suite written by the C-suite.

  • Media Blog

    A blog for and about the media featuring trends, tips, tools, media moves and more.

About Jim Dougherty

Jim Dougherty is a featured contributor to the Cision Blog and his own blog, leaderswest. His areas of interest include statistics, technology, and content marketing. When not writing, he is likely reading, running, playing guitar or being a dad. PRSA member. Find him on Twitter @jimdougherty.