Which tool to do Analytics?

Data

Business

UX

Analytics

Lean Startup

Product

This article is the second of a series of three articles aiming to answer to What is Analytics, Why, When and How to use it

In my previous article, we discussed why Analytics is a great mean to drive your product. We will now take a look at one of the most famous Analytics tool.

The purpose of this article is to give you a brief but nonetheless actionable knowledge of what you can do with Google Analytics (GA), and at first, why you would use GA.

Why Google Analytics?

Before even digging into the tool itself, let's detail why you would use GA instead of another solution.

On the bright side, GA is:

  • Easy to set-up: The basic set of features can be implemented in minutes (then data still needs to flow...)
  • Used everywhere: If you have a question, there is a high chance someone asked already
  • Full of advanced features: You can get insights about your audience (the profile of your customers), your acquisition (the performance of your online marketing campaigns), your users behavior and your goals (the main performance objectives of your app/business)
  • Free: All of the above points, for free, without real limitations
  • On mobile apps and websites: The tool tracks both types of platform, an additional feature lets you combine the data for cross-device analysis
  • Plugged to Google infrastructure: It gives further details about your customer profiles, thanks to what Google collected on them across their network
  • Well connected to marketing tools: GA is plugged to Google main online marketing solutions: It displays marketing campaign performances from those solutions (like Adwords), but also sends to those tools customer behavior data used for targeting
  • Paired with Google Tag Manager (GTM): A tag management system allows any non-tech person to quickly understand and change the tracking of your platform without impacting the tech team, and it also redirects data to several marketing tools at once (Google, Facebook, Criteo, etc.)
  • Including a Premium (paying) version: GA Premium gives more advanced solutions (like querying directly in the time-series database of GA through BigQuery), this option existing is a plus, but ...

On the dark side, GA is:

  • Including a Premium (paying) version: ... from GA basic to Premium, the cost is high
  • Not on your servers, but Google's: You don't store your data privately, if you care about this, it can be a real blocker
  • Complex but shallow: It seems complex at first, but once you master it and want to dig more into your data, you'll realize GA dashboards don't give enough insights, and that's where GA premium comes in...

For most Analytics needs, GA is a terrific solution, it has many more pros than cons. And I personally chose to use it with success many times.

However if you feel that something is missing or wrong, or that a new challenger entered the scene, please tell us in the comments section!

What can you find in GA?

dims-and-metrics.png

A dimension represents an attribute (a characteristic) of an entity, like:

  • The name, age or location of a user
  • The date of creation, size or file extension of a picture

Whereas a metric is an aggregated numeric value used for analyzing and comparing dimensions:

  • A number of users, sessions, pageviews or downloads

For a proper analysis, one does not come without the other, the dimension gives a definition and scope to your analysis, and the metric a point of comparison.

dims-and-metrics-ex.png

GA has many dimensions (and you can even create yours), categorized in the tool as:

  • Audience -> Users profile
  • Acquisition -> Users origin (thus online marketing performance)
  • Behavior
  • Conversions -> Platform goals

However as soon as you want to start refining your analysis, creating reports and calculating values, you will need to find external solutions to GA, since GA user interface does not include those features. And to do so, it is good to know how the data is stored in the database of GA.

time-series.png

Each time a user "interacts" (as long as the interaction is tracked), a hit is sent to GA and stored in a time-series database (DB). You can picture this type of DB as a huge and unique table where each "hit" takes up one line and where the columns are mainly dimensions (hence attributes of hits). Then Google aggregates (counts) by dimensions to feed the metrics of its pre-configured user interface.

What's more interesting is by handling this DB by yourself, you can very easily do your own aggregation with specific business rules in order to calculate and compare advanced complex KPIs.

In the following article, I tell you how to implement Google Analytics for React Native.