QUICKSIGHT

Amazon QuickSight is a fast, cloud-powered business intelligence service that delivers insights to organizations. QuickSight allows its users to create and publish interactive dashboards and machine-learning enabled insights into data. Dashboards can be accessed from any device and embedded into applications, portals, and websites.

MY ROLE

2020 was marked with heavy investments into enhancing Quick Sight’s usability and scaling its core interaction feature offering. The QuickSight organization was committed to enhancing QS’s dashboard interactivity features, allowing readers better filter and drill down experiences and bringing information to life for its reader user base. I was solving for both dashboard authoring workflows as well as dynamic dashboard features and the reader user experience.

ROLE

My team consisted of Product Designers, User Researchers, Design Technologists and Visual Designers. I defined a strategic product design direction for dynamic dashboards focusing on filtering and drill down interaction models. I worked cross-functionally to defining and align on development roadmap priorities. I was also responsible for supporting key launches for core analytics focusing on filtering, dashboard controls, parameter creation, and authoring workflows that enabled for a more streamlined reader dashboard CX.

I partnered closely with a researcher in conducting both generative and task-based studies to baseline existing tools across top tasks. Our success was measured as year-over-year improvement to SUS score of the tools, Voice of customer quarterly surveys as well as more targeted engagements with our user based.

IMPACT

We were working towards an increase in adoption of core dashboard interactivity features, as well as 10% improved usability measured through quarterly baselines.

The Problem

Core Dynamic features not supporting readers

Today, Authors are manually configuring filter actions on visuals in order to enable dashboard interactivity for readers. We know that 80% of filter actions are author configured on pivot tables (55%) and tables (24%) as the top two chart types configured for cross-visual interaction, followed by Donuts (6%), vertical bars (6%), horizontal bars (6%) and pie charts (5%). Authors choose a single-click interaction model for their readers (98%) and right-click navigation menu (2%) to allow cross-visual filtering today. We have heard authors needing more support to conduct on-screen data analysis as well as Readers wanting more flexibility to slice and dice data on the screen.