Use the Google Analytics Search Bar for Complex Queries in 2020

Google Analytics’ ‘Ask Analytics Intelligence’, enables users to ask questions in natural language to get results via the search bar. Let’s understand how you can use this powerful feature for crucial insights in 2020.

Businesses rely largely on Google Analytics (GA)Opens a new window for its powerful web reporting capabilities. However, despite its powerful reporting capabilities, organizations often struggle with deriving insights from GA data. The reason is obvious – with so much data, how do you drill down through vast data sets, and all those options, to get the exact information you need?

To resolve this issue, Google released Google Analytics Intelligence, a machine learning (ML) powered suite to help Analytics users get the exact insights they need. One such feature is Ask Analytics Intelligence that allows users to get key insights just by asking questions via the search bar. 

What Is Ask Analytics Intelligence?

Launched in 2017, Ask Analytics Intelligence is a feature that lets users get answers by typing natural language queries. For instance, a user can type in how many people came to the website from the US in 2019, and Google Analytics will provide an answer like this:
 

Before we decode how Ask Analytics Intelligence works, let’s take a step back and decipher the concepts of dimensions and metrics:

1. Dimension: Dimensions are the attributes of your website visitors. For example, gender, location, device, browser, etc.
2. Metric: Metrics are quantitative measurements. For example, sessions, bounce rate, goal completions, etc.
    

When a user types in a query, Google Analytics breaks it down and processes it as “dimension →  dimension value → metric → time period” to provide the answer. To process the query, Google uses the same natural language processing technology as Google Search, while machine learning (ML) is used to decide the accuracy of the result. 

Now that we know how it works, let’s look at how you can use this feature to get the answers to complex queries.

1. Ask ‘Why’ Questions

If you want to know the root cause of a certain behavior, you can frame a why question to get an answer. For instance, you can ask: Why did website visits decrease this month? GA will provide an answer, depicting the difference between this and the previous month.

The change exploration feature provides a plausible explanation for what caused the change. Some of the common dimensions that the tool considers to evaluate the change are user types, locations, and channel groupings.

Ask Analytics Intelligence provides answers to ‘why’ questions along with plausible explanations
 

Learn More: What Is Marketing Analytics?Opens a new window

2. Evaluate Performance

There are so many metrics to track for a website. You need to know the performance of various content types, traffic sources, product sales, ad campaigns, keywords, and so on. You can type in queries to evaluate the performance of these areas. For example, to identify top-selling products, an e-commerce store owner might ask: What are the top five selling products?

When asked: What were the top performing channels since June? GA returned the following result:

Users can ask questions to evaluate the performance of various website-related areas
 

3. Identify Trends

While performance measurement gives you a holistic picture (which is often a number), if you want to drill into the nitty-gritty of data, you need to analyze trends. Trends help you understand the performance of a certain metric over a time period. So, if we take the above example of top-performing channels to identify trends, we get this:
 

You can identify trends in the existing data
 

Learn More: Tech-Driven Marketing: Four Capabilities Marketers Need TodayOpens a new window

4. Detect Anomalies/Outliers

Anomalies are patterns in the data that do not belong to the expected behavior. They are reported if the actual result doesn’t match the expected result by a significant measure. You can use the Insights feature to identify such outliers in the data. Continuing with our previous example, we can see a sharp rise and fall in the referral traffic, and based on this information, we can ask is there a problem with referral traffic since June? to find anomalies:

5. Compare Data

If you want to see how different devices are contributing to the traffic or want to know which cities drive the most traffic, you can do so by asking comparative questions. For instance, you can compare the traffic for organic and direct as the medium to know which one is more effective:

You can compare two or more dimensions to get an understanding of their performance
 

Learn More: Using Technology to Gather Human Insights, Not Just DataOpens a new window

Creating Complex Queries

We’d recommend that you check out Google Analytics’ primer on the types of questions you can ask. Simply click on the search bar in Google Analytics followed by What questions can I ask? from the Insights section.

How to access the “What question can I ask” feature of GA

You can build on the examples we showcased to get the information you need. To get crucial insights from the data, you need to know how to frame queries effectively. Keep the language as simple as possible when framing questions. For instance, if you want to compare the percentage of visitors from the United States and Canada during 2018 and 2019, you can ask frame the query as: How did the share of users compare in 2019 vs 2018 from the US and Canada?

Closing Thoughts

To get accurate results, be clear while forming queries, especially when viewing comparisons. For instance, if you use “and” instead of “versus”, you’ll get a combined result of both. As with any AI/ML-driven product, Google Analytics Intelligence isn’t completely developed yet. So, whenever you get an inaccurate result, make sure to provide feedback to make it better.

Do you have any questions on how use the Google Analytics search bar for complex queries? Ask us on TwitterOpens a new window , LinkedInOpens a new window , or FacebookOpens a new window .