By Berit Hoffmann - December 17, 2020
This post was originally posted on the Looker blog on October 9, 2020.
Today, everyone needs the ability to make data-driven decisions quickly. To meet this increased demand for data accessibility, organizations are breaking down barriers to insights by using platforms like Looker to embed dashboards into their teams’ existing workflows.
With the level of data granularity and access users can achieve with Looker, it makes sense that the number of questions like “why is my KPI going up (or down) this week?” will increase.
Herein lies the challenge.
Even with a healthy data model, answering why-questions can require hours of manual work in the data. By the time all the necessary work is done and the data is examined to answer the initial question, new data has already been created — thus resulting in an incomplete answer.
Fortunately, teams can bridge this gap between data questions and answers by combining the technologies of Sisu and Looker. With the Sisu platform, you can automatically explore all the factors — and combinations of factors — that may be influencing changes to your metrics and quickly surface answers on the subpopulations in the data that are having the greatest impact.
Imagine you’re a data analyst at a large, online streaming service — an industry where the competition for viewer time and attention is fierce. To stay competitive and successful, it’s critical that everyone has access to the business’ real-time streaming analytics.
As an analyst in this industry, the average dataset you’re likely working with on a daily basis includes hundreds of columns, with data spanning across categories like household demographics, content viewed, and total session minutes. This adds up to billions of data points that, when analyzed, can reveal trends or new business opportunities. However, with so much data to work with, it can be difficult to know where to get started.
Augmented data workflows enable analysts in situations like these to speed up the time it takes to get from data processing to actionable data insights. And because these workflows are based on the metrics that are core to the business, Augmented Analytics enable you as an analyst to go from being a data bottleneck to facilitating rapid data discovery across teams in your organization.
So going back to our example, if you’re an analyst working in Looker, you can get started with Augmented Analytics by simply connecting Sisu to the data you’ve already prepared for use in Looker. Sisu seamlessly integrates to your Looker dashboard via a direct link that connects your dashboard to the corresponding KPI analysis in Sisu. By eliminating the need to manually manipulate data, your time as an analyst can now be focused on new business projects or initiatives, rather than time-intensive processes and ad-hoc requests.
With Augmented Analytics a part of your workflow, you can more quickly analyze vast amounts of data and begin making observations and decisions based on real-time insights.
For example, let’s say you’re looking at the Looker dashboard below, all about viewer sessions. You want to use this data to specifically identify what subpopulations are having the greatest effect on overall session duration — an important KPI for your organization.
By clicking the link that takes you from your Looker dashboard to your analysis in Sisu, you can see that overall session duration went down by 2%. And with Sisu, you can further investigate how different subpopulations in the data are influencing the overall session duration metric. In the platform you’re able to see that viewers in Canada watching Stranger Things have significantly increased their average session duration over the last month, going from less than 40 minutes to over 60, which then positively impacted the overall session duration metric.
From here, you can continue to filter on various data factors and build upon this initial insight, all from within Sisu and without needing to rerun the analysis. By using Sisu and Looker to prioritize and dive deeper into subpopulations in your data, you can quickly uncover insights and use your findings to inform decis