Posted by Lois Haggard · Feb 19, 2019 at 10:52 PM

Web Dashboards

New Mexico plans to develop a web dashboard as part of our project. We noticed that a few other team projects also involved dashboards. The those teams, I'd like to get your input:

The term, "dashboard," can mean a variety of different things. What does the term dashboard mean to you? Are there quintessential elements of a dashboard?

Thanks,

Lois

Comments & Events

Tiffany Nicole Tsukuda

Hi Lois,

Our team in San Mateo County, CA will be developing a dashboard prototype as part of our efforts to update our current daily situational awareness/surveillance report.

In addition to doing some pre-dev work to define the purpose/scope of the dashboard, and plan which frameworks/technologies to use (e.g., data viz/dashboard products: Esri, Power BI, SAS Visual Analytics, Tableau; front-end custom programming solutions: Angular, Bootstrap, Chartj.js, Kendo UI, R Studio Shiny, etc.), here are some of the considerations that we have compiled/will be looking at:

Data platform vs. dashboard:
- Data platform: End-to-end BI platform/suite with integration/ETL tools, database/data store, and self-service tools for reporting and analytics (e.g., dashboards); data can be cut into many different dimensions and views.
- Dashboard: Interactive web app (details below).
Dashboard type (operational vs strategic vs analytical):
- Analytical:
  - Functions: Investigate trends, predict outcomes, discover insights.
  - Main goals: Help users make sense of the data, generate ideas for future analyses, and make decisions.
- Operational:
  - Functions: Monitor performance of a process; track performance of key metrics, indicators, and resources.
  - Main goals: Highlight deviations and issues that require attention; show status of critical information to help users be proactive, efficient, and agile.
- Strategic:
  - Functions: Monitor the processes that support strategic initiatives, measure progress towards strategic goals.
  - Main goal: Measure the current state against the intended outcome/future state.
Target users: General public, department, specific stakeholder group, leadership, etc.
Guided informative vs. freeform exploratory: Control the data narrative with a set point of view, or create a freeform exploratory experience that gives users the ability to drill down and interrogate details.
Personalization: Create specific content, experience, and functionality based on a user’s identity/group permissions (system-driven).
Customization: Users can configure the dashboard layout, content, or system functionality, e.g., pick and choose which data to display (show/hide panels/widgets) and how to display it (configure row/column layout, re-order panels/widgets, expand/minimize panels), and apply a customized theme (change colors, fonts, font sizes).
Parameterization: Users can filter, sort, select ranges, group, etc.
Representing the data:
- Determine which visualization or object (e.g., chart, graph, grid/table, map, timeline, narrative, photo, video, graphic, etc.) to use to represent the view of the data (e.g., relationship, comparison, composition, distribution, change over time, etc.).
- Viewing patterns: Ratio of quantitative vs qualitative data.
Other considerations:
- UX/UI best practices and design standards, Gutenberg diagram
- Section 508 web accessibility standards
- Responsive and cross-browser compatible
- Refresh rates: Real-time/near real-time vs daily, weekly, monthly, quarterly, annually, etc.
- Enable other ways to consume the data: download CSV or Excel file, print PDF report view, provide API URL, etc.

Design, functionality, and quality elements/principles:

Embrace minimalism: No more than 5-9 visualizations per page, data are revealed as needed.
Design an intuitive and informative experience: The default view of the content is easy to search/navigate, has the functionality to conform to the experience level of the user, and can provide the appropriate context and information to optimize the user’s experience.
Data tells a clear story: The purpose of the dashboard and how to use it is well-defined, and the chosen data visualizations correctly represent the data and the information that can be extracted from it. Clear/consistent naming conventions, definitions, and formatting are used.
Data are grouped logically (some examples below):
- Inverted pyramid: Leading/noteworthy indicators on top, trends and important details in the middle, general background info/granular details at the bottom.
- Analytic pyramid: Supporting data on top, conclusions from data on bottom.
- Squix pyramid: Actions on top; comparisons, trends, and KPIs in the middle; data at bottom.