How to Analyze User Behavior with a Google Website Survey

Maybe the title of this post is a bit over the top. But my goal is to show you how the four simple questions from a Google Website Survey can provide deep insights you can’t get from Google Analytics and most quantitative tools.

The four questions on the survey are:

• Overall, how satisfied are you with this website?
• What, if anything, do you find frustrating or unappealing about this website?
• What is your main reason for visiting this website today?
• Did you successfully complete your main reason for visiting this website today?

These four questions will help you understand the intent of your site users. Why are they visiting? What are they really trying to do?

In my last post I wrote about how to set up a run a Google Consumer Survey on your site. Now let’s take a look at some of the data.

Using the Google Website Survey Tool

To look at the data for a survey first choose your survey, then select a question. You’ll notice that GCS actually provide a statistical significance calculation and let you know if there were enough answers. Consumer Surveys is looking for at least 95% statistical significance.

Choose a question from your website survey to analyze.

Let’s take a second to get acclimated to the reporting interface. On the left hand side there are a number of segmentation options. You can segment based on demographic data: age, gender, location, and income. Some of the segmentation data is inferred.

Inferred data is based on IP addresses. Income, and urban density are approximated using US census data. You can actually check out your own inferred data using your Google Ads Preferences.

There are lots of different features for the survey tool. You can segment data using the tools on the left or export each individual answer.

Other report features include the ability to navigate between questions (in blue) and the ability to export the data (in red, see below).

Exporting is really cool. Not only does it include all of the segmentation data for each response, it also includes the page on the site where the user answered the question.

You can export each survey. The exported CSV file will contain each question along with demographic attributes of the user.

Finally, The data for the question, ie the user feedback, is located in the center of the screen. The format of this data will change depending on the type of question.

Here we’re looking at a multiple choice question. We can see the most people are satisfied with the site.

You can see the number of user responses for each question, along with a statistical significance calculation. The survey tool is looking for 95% confidence before declaring a winner.

Drawing Insights from the Data

Now let’s walk through the questions and talk about how useful they are.

“Overall, how satisfied are you with this website?”

Site satisfaction is a great metric. It’s a good way to gauge overall sentiment around your web experience. That’s why there is a trend graph at the top of the data.

Site satisfaction is a great way to understand the overall happiness of your users.

The reason this metric is important is that it’s not aligned to any general goal. It applies to this user and their specific experience. Something you can not get from quantitative data.

Let’s talk about segmentation for a minute. While the top-line satisfaction metric is great, we need to segment to understand why. Use the segmentation tools to segment each answer by your desired demographic.

Is there a particular gender that is driving the satisfaction? Or age-group. The segmentation is also very useful if you are trying to understand if your perceived audience. Are these the people you were planning to attract?

It’s important to segment your survey results. Start by segmenting by gender and geographic location.

Are you attracting and satisfying your core audience? I certainly hope so! If not, you may need to change your overall marketing strategy to attract a different audience.

Now let’s move on to the second question.

“What, if anything, do you find frustrating or unappealing about this website?”

Open ended questions are great. You get the raw opinion of your users. There’s nothing like an f-bomb to make your team squirm :)

You’ll notice that the results looks a bit different. The tool has included a word cloud to make it easier to view common terms and feedback. I should mention that the tools does some manipulation of the data. It forces the answers to lowercase to eliminate duplicates.

The Google Website Survey tool will use a word cloud to help you understand free-form text entries.

Working with free-form text can be hard due to synonyms and misspellings. Consumer Surveys tries to group similar terms together. You can also customize the groupings by dragging and dropping similar terms together.

You can group similar terms together, making it easier to understand user feedback.

This is SO helpful to make the data more actionable. Free form data can have numerous variations of the same things.

The third question is actually my favorite.

“What is your main reason for visiting this website today?”

This is all about visitor intent. Why did you come here today? That’s something that’s really hard to discern from an inbound keyword, landing page or campaign information.

The word cloud can be very helpful here. It’s easy to pick out common words that represent what people are interested in. In this data, from a food site, you can see that most users are interested in finding recipes.

Understand _why_ visitors are on your site using the word cloud.

Obviously a lot of people are interested in recipes :) Notice all the different variation?

Some of you may have observed, that this data can be VERY valuable for search engine marketing. Look at all those beautiful keywords. Are you bidding on them? Are you optimized for them?

Here I might segment by gender to see if men are interested in different recipes than women. If they are then perhaps I change my marketing strategy and try to connect with each audience using the types of recipes they are interested in.

The last question is a single answer question that asks the user if they were able to complete their objective.

“Did you successfully complete your main reason for visiting this website today?”

This question helps us understand if we’re meeting the needs of our users, regardless of their reason for visiting. If we compare this to the quantitative world it’s like asking the user if they converted or completed their goal.

The nice thing about this data is that it’s adjusted for this specific user’s intent – “… did you successfully complete your main reason for visiting…” This is very different than just a generic conversion rate.

This is another great metric to trend over time. If fact, the survey tool will trend this data, one point for each month. Again, this is great data to add to a dashboard.

Google Consumer Survey will trend task completion rate over time. One data point for each month.

Don’t forget to segment the results. Gender, geographic location, etc. All very important.

While the quantitative data we get from Google Analytics is great, it’s the qualitative data that really helps us understand why things happen. We can use this data to discern intent and then develop new ways to be successful.

Are you using a Google Consumer Survey on your website? Have you gotten any interesting insights? I’d love to hear about them. Leave a comment below!