3 Fundamentals To Know About Google Analytics Before Doing Analysis

What Can You Get From This Article?

  1. Understand that Google Analytics collects and aggregates data on many levels, or “units of analyses”, and be able to identify the correct level for your analysis.
  2. Understand the key differences between metrics and dimensions and be able to pair correct metrics with dimensions in order to answer different business questions.
  3. Know how to choose meaningful metrics for your analysis that can not only help you answer business questions but also can drive action.

In the first blog post of our Google Analytics series, we outlined 4 business questions Google Analytics can answer (link here). Now, it’s time to talk about how we can answer those questions.

However, before we dig into the questions, it’s important that you familiarize yourself with 3 fundamental principles that are crucial to pay attention to when using Google Analytics.

Talking about these 3 principles will not only give you the context you need to help you better understand the four questions, it will also help you both answer important business questions and drive the actions.

Let’s begin!


Fundamental 1: What kind of data is Google Analytics collecting? What is the “unit of analysis”?

Why is this important?

The unit of analysis is like a guiding light that can provide a structured way of understanding where to look for the data you need when formulating business questions.

For example, when approaching typical web analytics questions like “Where is my website’s traffic coming from?”, you should first breakdown the question into two levels — the session level and the user level.

Let’s start with the session level. Google analytics can tell you specifically which channel (such as search, social media, etc) your session visits are coming from and what visitors’ behavior is like from those individual sources.

Understanding session level data can help you optimize visitor experience across the different channels.

However, your analysis should not stop here (as many people do). Rather, in order to understand where your traffic is coming from by session you need to analyze your traffic sources by user.

When analyzing your traffic sources by user, your question becomes “How are different groups of users interacting with my site?”

For example, Google Analytics can help you find out if users between 25–34 years old first encountered your site through social media or another source. Taking this a step further, it can also show you whether or not they eventually made a purchase on your site.

Analyzing traffic sources on a user level can help you take a user centric approach and thus, in addition to optimizing your visitor experiences across different traffic sources you can also design specific visitor experiences for your core audience groups. This will further improve their likelihood of conversion.

The Details

Let’s begin our exploration of the different units of analysis by talking briefly about how Google Analytics collects data.

On a very high level, data collected by Google Analytics can be divided into four descending levels (also called “units of analysis): visitors, sessions, page views, and events.

To explain what these four levels entail, let’s welcome back our friend Patrick — your site visitor who was the narrator of our last blog post:

  • Visitor Level: Every time Patrick visits your website, Google identifies and collects demographic information about Patrick such as his age, gender, and interests.
  • Session Level: Patrick might visit your websites multiple times and each time he visits he will start a “session”. GA collects data about where this specific “session” took place, the length of the session, and whether the session came from Google Search, social media, or another source (this is also called the channel).
  • Pageview Level: During each “session” Patrick might visit multiple pages on your website. Each page visit is called a “pageview”. GA collects data about what order he visits the pages and how long he spends on each of them.
  • Event Level: Finally, when visiting each page Patrick might engage in actions such as clicking a button or watching a video. These actions are called “events”. Google can track the occurrence of events and calculate how frequently these events are happening.

Logistically, Google Analytics supports data collection on visitors, sessions, and pageviews without any additional setup. However, in order to track events you need to embed additional code into your website.


Fundamental 2: What is the difference between metrics and dimensions and what kind of business questions can they help me answer?

Why is this important?

Understanding metrics will help you gain clarity about what kind of information Google Analytics is collecting. However, tracking metrics without dimensions can only give you a broad overview of the information being collected.

If all metrics are performing well, it’s difficult to isolate what’s really driving the positive change; an important step if you want to take advantage of the momentum. Likewise, if some metrics are not performing well it is difficult to tell what exactly is wrong.

Even worse, metrics may present well but in reality be hiding underlying issues.

For example, you may be losing all of your small customers while gaining one large customer. As a result, your revenue metric maintains or even improves., Thus, you’re left in the dark about your fleeing small customers.

This is exactly why you need to understand how to slice and dice your data into multiple dimensions. It will allow you to not only test your “known unknowns,” but also to discover your “unknown unknowns” — the hidden drivers of positive or negative trends that you didn’t even know to look for.

The Details

There are two types of variables in Google Analytics: metrics and dimensions.

Metrics are quantitative measurements and are collected by Google Analytics at various levels. These tell a story about user experience on your website. Common metrics include:

  • Total Sessions: The number of sessions a visitor took part in on your website. A session expires if a visitor does not perform an action for 30 minutes. A new session begins if he/she resume action.
  • Bounce Rate: The proportion of users that visited your landing page and then left without accessing any other pages.
  • Session Duration/PageView Duration: The average length of time visitors spend on a specific page or partaking in a session.
  • Pages/Session: The average number of pages visitors visit each session.
  • Conversion Rate (requires setup): The proportion of visitors that successfully perform “conversion actions.” This could be filling out a lead form or making a purchase. This requires goal setup.
  • Total Revenue (requires setup): The total number of purchases made by visitors to your website. Requires e-commerce integration.

If we were to just look at metrics, we would be able to answer a lot of business questions about the overall performance of your site. Here are the questions you can answer with various metrics.

  • Are people visiting my website at all? (total sessions and users)
  • Are people coming back to my website? (number of new vs returning users, sessions per user)
  • Are people going beyond my home page and looking at other pages? (bounce rate)
  • On average, how long are people spending on my website? (average session duration)
  • Overall, is my website accomplishing my business objective; whether that is selling my product or collecting leads? (conversion rate)

However, only analyzing metrics on a general level cannot produce many insights for users. In order to remedy this, GA provides many more attributes that help analysts slice and dice the data. These attributes are called dimensions.

Common dimensions in Google Analytics include:

  • Demographics: Includes the attributes of a user such as his/her age, gender, and interests.
  • Location: The geographic location the website is accessed from including everything from the city to the continent.
  • Behavior: Whether the users are new or returning and how engaged they are with your website in terms of repeated visits and session length.
  • Devices: The devices the sessions were conducted on; including Mobile, Desktop, or Tablet.
  • Channels: The channel source of those visits; including Direct, Referral, Social Media, Organic Search, etc.
  • Pages: The pages accessed during the visit.

Fundamental 3: How do I choose metrics that drive business decisions and actions?

Why is this important?

Good analysis can tell you how to take advantage of opportunities. However, it can’t give you real time updates on whether or not those actions are having the intended effect.

Metrics condense those analyses into few key focal points that you can monitor in real time in order to see the effects of your actions. This will allow you to see whether your actions are effectively tackling your challenges and taking advance of your opportunities.

Because metrics are meant to be agile and actionable you shouldn’t track too many metrics at any given time (around 10 is usually a good number).

Therefore, be careful when you are choosing which ones to track. One bad metric can be detrimental to your implementation process.

On the other hand, a good metric will not only accelerate your implementation process, it will also offer your team a clear goal to pursue.

The Details

It is important to note that, depending on the area of your web presence that you want to focus on, the metrics you should choose will vary greatly.

That’s why in this session we are going to offer a general guideline on choosing the correct metrics . We will touch on more specific metrics when we dig into each of the analytics questions.

According to Alistair Croll & Benjamin Yoskovitz in their book Lean Analytics, a good metric has the following attributes:

  • It is comparative — This helps you understand where you are compared with your competitors and past performance. For example, you can see if this metric has gone up, down, or stayed the same during the last 3 years. Likewise, you can see if this is an industry-wide issue or a company-specific issue.
  • It is understandable — This helps you clearly communicate it with your entire team and gain buy-ins.
  • It is relevant enough to drive action — This will push you to take real action based on your analysis.
  • It is a ratio or a rate — This makes it easily comparable and can capture multiple pieces of information at the same time.

On top of those four insights, I would also like to add two points specific to Google Analytics:

  • It is granular — This helps you capture underlying trends that impact the users’ experiences on your site.
  • It is general (to a degree)-This helps prevent paralysis that can sometimes occur when you’re overloaded with information. Also, if a metric is too granular, it might be capturing “noise” instead of “signals. This is confusing and less than ideal.

When choosing metrics, it is also important to avoid what people call “Vanity metrics”.

“Vanity metrics” are metrics that do without context. . For example, a classical vanity metric is the total sessions for your site during the past week.

Without comparison, it is impossible to tell whether this number is good or bad. 10,000 visitors per day would be good news if you only had 5,000 the previous month but it would certainly be bad news if you had 15,000.

To avoid vanity metrics, always analyze your metrics in context by comparing them either internally with your past performance or externally with your competitors.

This will not only help you better understand the current status of your website, it will also help you come up with more actionable goals and objectives to move towards.


Thank you for reading this blog! Next week we are going to dig into actual business questions by exploring how Google Analytics can help you understand who your customers are.

We will also release the first version of our Google Analytics tool shortly after the next post. Stay tuned!

To stay updated on future blog posts, please follow us on Medium at analytics-for-humans, or on Twitter and Facebook. If you have any questions about this article, please feel free to email me at bill@humanlytics.co.

This blog is produced by Humanlytics. At Humanlytics, we are making tools to make Data Analytics easy, compelling, and valuable for all businesses. If you want to learn more about Humanlytics, please visit our site at humanlytics.co.

Show your support

Clapping shows how much you appreciated Bill Su’s story.