So you are tracking your site or app in Google Analytics (or Adobe or Webtrends or AT Internet or Webtrekk or Piwik and sundry). Assuming you configured your tracking correctly, you are now enjoying reports in a user-friendly interface. Hurray for data!
However, the reporting interface only goes so far and will eventually generate frustration for you as a digital marketer because you need to use the raw data behind the reports. In this post, we’re going to list ways to extract and exploit data from your favorite analytics platform in ways other than exporting plain PDFs. The various techniques become increasingly sophisticated as we progress down this post.
Depending on your organisation’s analytics maturity level, you will be sharing insights with your team and/or your+1’s and their +1’s. So you need to provide said insights in a format they will understand. Just sending out a PDF curves and graphs is not going to cut it. Page views and sessions are great vanity metrics but what kind of story do they tell? What does a geo/map report say about the state of conversion by segments?
That’s right, you’ll need to get that data out of the platform and rework it somehow to display it in ways that allow you to effectively work with it and in a format that your higher ups will understand.
Low-hanging fruit: using built-in export functions
In the Google Analytics interface, just about any report sports an “Export” button above the graphs and data tables. From there you can select a handful of export formats.
Pros: looks “pretty”; easy to view across platforms; somewhat searchable; easy to print at scale
Cons: rendered format, meaning it cannot be modified; data can’t be extracted
Pros: Actual numbers you can use and import into your favorite datacrunching solutions, although many of them have now developed APIs that connect go Google Analytics’.
Cons: Ugly as heck. Then again, they’re just numbers 🙂
Pros: Used by entire industries to share data; data can be integrated into other sheets; formulae can help
Cons: Excel is still very clunky, despite dozens of versions. Number formatting can be hell for international (non-US) users.
Pros: Leverage the Google ecosystem; awesome sharing/collaborative capabilities; *almost* functionally comparable to Excel
Cons:*almost* functionally comparable to Excel; hard to access in Google-unfriendly companies and environments.
Great, now you know that you can easily extract data from Google Analytics, in enough formats that you can actually use them on a daily basis. But as you can imagine, the goal of posts such as this one is ideally to broaden your horizons and show you alternatives.
Let’s move on to advanced methods!
Advanced extraction: using the Reporting API
(aka the golden apple)
An API (Application Programming Interface) is a way for developers you access your applications, query its data in both ‘read’ and ‘write’ mode.
The Reporting API arrived fairly late in Google Analytics’ evolution and still is catching up to the platform’s features but I for one use it on a daily basis, either using code I wrote/borrowed or using a visualisation platform such as Domo or Tableau Software or others.
The API allows you to pull just about any dimension/metric in Google Analytics, with somewhat limited sampling
R / scripts:
Pros: Leverage the power of the API and use the data crunching/processing functions inherent to said scripts (Python most notably) and R of course.
Cons: Requires coding as well as decent (read: advanced) knowledge of statistics
BigQuery (Premium/360 only):
BigQuery is in a subcategory of its own, as part of the Google Cloud Platform. It is reserved for Google Analytics 360 (née Premium) customers. Please hold comments about how much it costs, we know…!
Pros: API extraction on steroids; essentially SQL for the full Google Analytics data set, and then some; full granular data; no sampling
Cons: Requires Premium/360
Data visualization platforms:
(Domo, Tableau Software, Klipfolio, BIME et al.)
Pros: They create pretty graphs and dashboards that can help you make sense of your data
Cons: Expensive; dashboard sharing capabilities not always adapted to corporate policies
On the topic of data visualisation, give Google’s Data Studio a try, it will allow you to create fully customizable, shareable dashboards based on your Google Analytics data (available outside of the US “soon”).
So you listed export methods, which one do I choose?
Finding your favorite/most useful export method is not easy but it boils down to how much slicing and dicing you want to apply to the data before/after it is extracted from the interface or the API.
Long story short, you need to check for obvious parameters:
- Dimension/column formatting
- Segment management
- Export size (batch exports, paging and report offsets)
- Time management (time stamps, timezone offsets)
- Date range comparisons
As I mentioned before, your reporting sophistication will vary, based on the following criteria:
- How your KPIs are defined for your company
- Your mastery of segmentation techniques
- How the data you extract will tie with other data sources
- How mature your organization is to act on data
- How much time and resources you or your team are willing to spend out of your already busy work week to generate reporting
In this post, I listed ways for you to extract useful & insightful out of Google Analytics. Again, the techniques I mention here can for the most part apply to other analytics solutions but Google Analytics is very likely the solution with which you’ll have the most immediate or prolonged exposure. And before you start trolling in the comments: YES, using a premium solution such as Google Analytics 360 can provide more extraction and processing options such as BigQuery or Data Studio 360 but let’s assume that us mere mortals cannot all get our hands on such sophistication 😉
But what about you? How do you use your analytics data? Do you work on it outside of the reporting interface to conduct analyses? Let us know in the comments!