Skip to main content

Sisu 101

Data analysis helps inform your business decisions based on facts you learn from your data. Exploring, analyzing, and fully understanding these data-driven facts help you and your team make more informed decisions, enact better policies, and implement more productive practices that serve to effectively “move the needle” of your organization’s success in the right direction.

Sisu provides you with facts about subgroups in your data as well as the degree to which they impact your selected metric. Exploratory analyses might lead to further questions you want to answer. You can make adjustments to an analysis as needed and run it as many times as you want to iteratively analyze your data until you reach the answers that can proactively inform your business actions and decisions.

Data-Driven Decision Making

Hidden deep within your data are valuable insights that are difficult to surface on your own. For example, did you know that sales of chocolate-peanut butter granola bars dropped among women under 30 in your New York stores, but are on the rise among the same customer cohort in Los Angeles? The impact of that fact is affecting your Los Angeles stores quite heavily! Maybe it’s time to make a product offering change in Los Angeles, and promote the same product more heavily in the Northeast with targeted marketing to young women.

Making data-driven decisions often looks like this:

assets/Discover_Ask_Reveal_Iterate.png

Steps to Analyze in Sisu

First you define a metric, which is basically the question you want to answer with your data (e.g., Average Sales Order Value). You then collect all of your data (you can either upload a CSV file or connect to your data warehouse)... and Sisu takes it from there!

assets/Analysis_Process.png

Designing Your Data

In analyzing data, generally you start with a business question you want answered (i.e., What is affecting my average sales order value?). You then define a Metric that Sisu uses to develop the Analysis to answer that question. You can further define the Analysis by adding filters or writing a SQL query to overlay the data.

An ideal dataset must consider 3 things

  • Granularity - Is the data at the right grain for the metric definition? What does each row mean?
  • Enrichment - Is the data rich enough to contain the factors of interest?
  • Actionability - Are the factors present in the data actionable to the stakeholder? Are these actions under your control?

The Sisu Framework

Sisu helps you organize your Analyses so you can run and maintain multiple projects, metrics, and results.

assets/Sisu_Framework.png

You’ll begin with a Project, which is a collection/folder of metrics. You can create multiple Projects.

Within each Project, you’ll define one or more Metrics. Sisu is metric-centric, which means you can configure and run multiple different Analyses within each Metric.

Sisu offers two basic ways to analyze your data:  through explorations and key driver analyses (KDAs). There are three types of KDAs: General Performance, Time Comparison, and Group Comparison. Explorations are a great place to start, since they provide a way to explore what is happening with your data. KDAs help you understand why your data is changing by providing insights on what is driving performance.