People have long used visualizations to reduce foreign or complex ideas to simple, consumable patterns. From ancient maps to the modern whiteboard, we gravitate to charts and pictures to communicate ideas. They allow us to quickly understand trends, create a jumping-off point for problem solving, and communicate conclusions.
But what happens when visualization becomes a thing in and of itself?
It is rare that aggregated charts and pictures of historical trends reveal root causes or uncover new opportunities. In fact, they hide the important details by design, and in so doing, prevent us from identifying and making precise and impactful changes. In the modern data cloud, comprised of billions of records and disparate data sources, visualizations are an important gateway to answers, but we now have more opportunity than ever to look to granular data to find the exact causes and relationships that can result in meaningful change. Sigma was built to take advantage of the modern data stack, and through our product, we have come to believe in a new set of data visualization principles.
Data viz should never be separated from the data itself.
For many years, analytics has been too focused on building visuals as a perfected end state. When dashboards are referred to as “permanent”, executives are treated to colorful and dense pictures, and, like the Emperor's New Clothes, convinced that they are informed. But modern data technology must allow us to freely move from those visuals to the most granular data and back, all while delivering great aesthetics with the performance and usability required to handle the most complex tasks.
Humans cannot make sense of, or get insight from, a billion-row table. Visualizations thus represent the view at the top of a haystack. The goal is then to follow that data down through the layers directly to the needle itself. A dashboard helps us monitor at a glance and alerts us when there’s a problem. But the ability to freely dive into the underlying data, without preset filters or other constraints, is critical to deeper analysis and problem solving. In other words, visualization becomes part of how you use data, but it’s rarely the goal itself.
In Sigma, users have the freedom to surpass the limitations of the provided workbook and answer their own follow-up questions without advanced analytical or data visualization skills. They can drill into any value on a chart without anyone having to pre-architect the drill paths for them. They can also expand any visualization into a spreadsheet view, add their own calculated columns, do their own groupings, and filter or sort to isolate the answers they are looking for. They can also create their own charts while using the original analysis from the provided ones. That means users aren’t building from the ground up, but taking what was given and digging their own way to their own needle.
Data visualization should bring clarity, not confusion.
Sigma’s visualization strategy brings forward the best of clarity and aesthetics with cloud-scale interactivity and granularity. We believe that visualizations should be beautiful; layouts should be responsive, colors should pop off the screen. And in our device-driven world, formatting should translate easily from your phone to your tablet to your monitor.
But, unique to Sigma, all of this should be easy to build by the average user, and scale seamlessly. This is because much of the work we do is ephemeral; it is the process before the conclusion. This requires fast iteration, collaboration, and communication. The ability to move seamlessly between scenarios and a chart allows teams to guide analysis effectively. And the lower the friction, the more stream of consciousness questions that can be asked and explored.
Sigma visualizations already apply the best practice chart layouts and formatting that are responsive to your screen by default. You can customize from there, but if you are not design-savvy, Sigma also provides color themes and shortcuts for different design configurations so that you’re not overwhelmed by unending (and unusable) options. Sigma’s KPI chart is a great example of a beautiful out-of-the-box element that can easily include advanced customizations like an area or trendline with a few simple clicks.
Even the layout of a Sigma workbook is defined with a similar experience to modern website builders that made web design accessible for non-designers. This means you can spend your time finding new insights instead of spending it making pixel-by-pixel modifications.
If you can’t take action on your data, your data analysis has failed. Being able to answer “Why?” and go from a pretty picture to action is no longer extra credit. That means you shouldn’t need to be a data scientist to figure out which action to take. Unfortunately, a great deal of data visualization today involves using data to create a work of art, as opposed to using that art to communicate action.
Everyone in a company needs the ability to build the 80% “good enough” version quickly and to be able to move at the speed of market. You should also be able to create a data visualization so easily, you’re using your own visualization to learn and understand the data in your own hands. You then get to look down into billions of rows of cloud data in a self-service way and find transformational insights.
In short, it’s time to think differently about visualizations. While visual appeal is table stakes, the ability to dive deeply into big data, create interactive experiences, and to self-service are the biggest opportunities today in the modern stack. With Sigma, we never build visuals for visuals’ sake, but to guide the user to action. We believe this is next-generation data visualization.