If you've read the first two installments from our latest ebook "Behavior Analytics: Understand What's in Your Event Data, and You Can Read Customers Like They're an Open Book," then you now know what event data is and why it's so important to your business. The ebook goes in-depth into the importance of understanding your users' actions, not just what they say, to build the best product or service. In this section, you'll learn the fundamental components of behavioral analytics and how Interana delivers a comprehensive behavioral analytics solution that's smart enough for data scientists and easy enough for anyone to use. You can download the free eBook here. Enjoy!
Start making sense: look at your event data with a view to answering your questions
Now that you know what event data is, we can start to explore how to perform behavioral analytics using Interana to start asking questions about users.
For example, you might start by asking what happened on your e-commerce site in the run-up to some important date, like Valentine’s Day. Let’s say you ran a promotion for people who had been on an online dating service within the past year, so you create a cohort of users who qualified for that promotion. Looking to generate insights, you start segmenting that cohort further by attributes you picked up from the dating service, including age group, income, and region. To figure out whether any of those attributes correlate to actions that influence conversion, you create cohorts for each attribute and build a funnel to see how close each user came to conversion. You might use session analysis to see who actually came to your site, how long they stayed, how many pages they viewed, and which pages seemed to interest them most. You might create a new metric, for example, which captures the ratio of time the user spends on the dating site every week and compares it to how long he or she spends on your site.
This brief example shows how you might start exploring the deluge of event data and uncovering insights. But don’t be intimidated. While it sounds very complex, it can be quite easy when you have the right solution.
The fundamental components in the example above are cohorts, metrics, sessions, and funnels. They are the building blocks of behavior analytics. Interana has made using these components easy by eliminating complicated query languages and providing users with a visual, interactive, and intuitive interface. Everything is done using pre-built features and guided wizards, so questions are built using drop downs and typeaheads, not code. Interana retains all the flexibility and freedom complicated query languages afford; we just made asking questions more intuitive and accessible.
Behavioral questions may be simple or complex, but you don’t want your behavioral analytics solution choosing what questions you can ask. You want complete freedom to build queries that can answer any question, across any dimension. Since Interana doesn’t place limits on the number of filters or features that can be combined, you can build on each question to create increasingly complex queries. Drill down and cohort users in ever-more specific groups, apply custom metrics, and run the results through a funnel. Every time you get back an answer that seems interesting, you can explore it further, asking additional questions, all with a goal of creating an even more effective promotion next Valentine’s Day.
And don’t be surprised if you suddenly look up from your screen to see that you lost track of time because you were so fully engaged in asking questions and getting answers. That happens to Interana users when they are in the zone because asking questions and getting answers is a fast and iterative process. But before you get started, let’s dig deeper into each of these components.
Group similar users together in Cohorts
Cohorts help you explore the many ways different groups of users engage with and use your product or service. Use cohorts to segment groups of users and then go deeper to create sub-cohorts based on characteristics you choose, such as demographic information, channel, retention rates, conversion rates, or type of actions taken over time. Cohorts help you determine the impact of new ideas or features, see what sticks, and answer other usage questions.
Measure what matters to you with Metrics
To answer questions that relate to statistics, ratios, or measurements about users’ behavior, create your own precise, customizable metrics. Get creative: look at metrics that correlate exactly to your industry, users, and line of business. You can isolate in a metric the events your users create as they interact with your products and services, then apply ratio metrics to any unit of analysis you choose. Activity metrics enables you to track emerging trends and how they shift over time, and if you see something intriguing, you can always use additional metrics to dive into users’ actions per session.
Be a fly on the wall for the duration of each Session
Define a session—such as the period between login and logout—and you can see all the actions users take and all the events that occur during that time period. For example, you can look at how long people stay on a certain page, or how much time they spend logged in. You can also take a step back from what happens within a session, and take in the bigger picture of session frequency to analyze the amount of user activity by the week, day, hour, or month.
Find out who completes the path you’ve laid out—or where they step off—using Funnels
Funnels track the steps users take through a pre-determined flow and provide a visual representation of how users convert, get off-track, or drop off the path that you’ve laid out for them. They show you how users interact with your products and services, how long they stay in each step of a flow, and what happens between steps. You can explore sub-populations within a funnel or a combination of funnels to unearth granular information on how users change as they interact with your product or service and understand what influences your users to complete a flow. You can also monitor users over time with multi-variable funnels.