The Art of Asking Questions - Google Analytics Intelligence
How to ask the right questions with Google Analytics Intelligence
Google Analytics Intelligence (GAI) is an incredibly powerful tool that gives you extra insight into your business and customers. This article will provide you with an understanding of good questions to ask GAI, the limitations of this tool, and a better understanding of the dimensions and metrics used by GAI. You can think of GAI as your digital golden retriever for gathering information about your business. While it understands some commands don’t expect it to offer answers to complex questions.
With GAI, you are talking to a computer, not a person. Knowing how it will respond will greatly decrease frustration levels and improve the quality of your queries. The computer doesn’t understand nuance or have any reference. This means you must be extremely specific about what you want to know, so the computer retrieves the proper data.
At this point, the system is constrained by what you can ask, so understanding the limitations is critical to achieving great results.
Google Analytics Intelligence Limitations
GAI can answer questions about data while pinpointing trends and patterns, but it still has limitations. Here are some examples:
- Why and How questions won't work, unless it’s “How many…”
- GAI won't give strategic advice on what to do with the data.
- GAI won’t answer "general" questions, like what the weather is
Now that we have a better understanding of what to expect from GAI, let's learn how to construct questions.
Understanding What the Computer Reads with Basic Questions
First, let’s look at a list of basic questions to ask the computer and highlight the keywords it uses to determine what answers you would like. Here are a few examples:
- How many users did we get yesterday?
- Where is my traffic coming from?
- How many new users did we get last week on mobile?
- Which countries are the most new users from?
- Which locations are new customers from?
Basic questions start with how, where, or which. For instance, you can ask, "How many new users were there this week" or “Where is my traffic coming from?"
Looking at the various questions you can ask, you'll notice the way the computer processes questions: it handles specific questions well. The more specific you are, the better data you receive.
Google e-commerce recommendation
For e-commerce, Google created 3 example commands. Like this:
- "Which products had over 200 unique purchases."
- "What is our conversion rate in Spain."
- "Share of revenue by country last quarter."
These are only templates, so you can change the questions to suit your needs better.
Dimensions vs Metrics - Variables Google Analytics Intelligence Understand
Some keywords such as "where" or "how many", represent locations or a numerical answer. You will notice how GAI picks up keywords such as mobile, new users, new customers, countries as well.
Google defines these as Dimensions.
Google defines their reporting variables as Metrics.
These keywords, along with the question structure, help GAI display their queries to you.
At the bottom of this article, you can see a list of Dimensions and a list of Metrics.
You can assess performance with questions starting with "which." Performance questions look like this, "Which cities had the most traffic." These questions are great for determining which locations are most interested in your services.
Trends start with words like ‘trend’ and a date command (Ex. daily, weekly, monthly). These commands will create graphs allowing you to visualize the data over time. We prefer to look at metrics on a weekly trend by using commands like, “Most traffic by week.” The data will show as year followed by the week number, like “201811” which is the 11th week of 2018.
You can use the ‘vs.’ command to have data appear side-by-side. For instance, saying “mobile users vs desktop users” will tell you the numbers for both sides next to each other. You can even compare trends like traffic with a command such as, “traffic this month vs last month.”
Share / Percent
The share or percent keyword will show the data from your request as a percentage rather than a total number. These keywords are useful when using comparison searches. An example of a percent command useful for understanding your customer base would be “percent of users by age.”
The above commands can be combined to find valuable data. Usually, ‘vs.’ is used to connect two pieces of data. Test some of the above commands yourself to find useful combinations for your business data. One potentially useful command would be, “percent of visitors by age vs percent of visitors by revenue.”
Using GAI properly can greatly improve your understanding of your business and customers. This incredibly powerful business intelligence tool is at your fingertips.
The GAI platform gains new intelligence each day, so lookout for articles detailing new features as they are released!
Dimensions for Google Analytics Intelligence
Here are a list of different types of basic level dimensions:
- Los Angeles
- New York
- Device Category
- 18 - 24
- 25 - 34
- 35 - 44
- 45 - 54
- 55 - 64
- 65 +
- User Type
- New User
- Returning User
- Affinity Categories
- In-Market Segments
- Traffic Sources
- Day of Week
Metrics for Google Analytics Intelligence
Here are a list of different types of basic level metrics:
- Items Per Purchase
- Transaction Per Session
- Total Refunds
- Refund Amount
- Added to Cart
- Product Removed from Cart
- LTV ( Cohert Enabled )
- Revenue Per User
- Pageview Per User
- Bounce Rate
- Session Duration
- Avg Session Duration
- pages / session
- Pages Tracking
- Entrance Rate
- Page Views
- Unique Views
- Average Time on Page
- Exit Rate
- Organic Searches
- Site Speed
- Page Load Time
- Avg Page Load Time
- Sessions Per User
- Weekly Users
- Monthly Users
- 1 Day Users
- % of New Sessions
- Goal Conversions
- Total Completed
- Abandoned Goals
- Abandon Rate
- Event Tracking
- Total Events
- Unique Events
- Event Value