Anatomy of a Modern Dashboard

5 Proven Approaches to Succeed

Modern dashboards contain summarized and interactive visualizations, and when done correctly, should effectively communicate performance. This guiding principle hasn’t changed in the last 10 years as the popularity of dashboards and data visualization soared. However, the technology choices, features, and approaches have advanced. Mobility, cloud and data exploration technologies have helped expand the persons who can create and consume dashboards.

People Before Features

A dashboard initiative that starts with data and tools, is likely to end in failure. This article won’t tell you how to plan or implement a dashboard, but before analyzing the anatomy of a dashboard, you should evaluate your dashboard approach with answers to the following questions:

WHO will consume information in dashboards and who will create them?
WHY those personas need a dashboard?
WHAT action or outcomes occur from usage of the dashboard?
WHERE will they access the dashboard (desktop, mobile, inside of CRM, etc)?
WHEN do you need the dashboard? Is this an urgent, short-term initiative or a long-term strategic solution?
HOW does the dashboard and its contents support your key organizational objectives?

The Ultimate Dashboard Approach

The ultimate dashboard tool / platform does not exist yet. However, there are great options based on your needs and available resources (data, skills, tools). At a glance, it is nearly impossible to tell the difference between dashboard tools because they all feature the same visual elements and features. Contrary to belief, your business users and information consumers won’t care what tools you use if you deliver a quality solution that addresses the questions listed above.

The following is a list of modern approaches to create enterprise dashboards:

1. Component-Based Dashboard Design

Component-based dashboards provide a library of visualizations, input controls, and data connections / objects which can be composed into an interactive dashboard apps. These ad-hoc dashboards allow users to filter, drill, rank, search and perform other tasks similar to reporting and data exploration, but with pre-defined and constrained data specific to business objectives / tasks. Modern ad-hoc dashboard platforms can include advanced capabilities for connecting visuals and granular control over data, business logic, and visual design.

Image Source: Antivia DecisonPoint

PROS CONS
Component-based dashboards provide a level of “self-service” exploration, allowing business users to modify results. The result of many ad-hoc dashboards produce similar results as traditional enterprise reporting. Additionally, some platforms require a deeper level of planning and specialized skills to ensure success.


2. Report / Exploration Summary Dashboards

A report or exploration dashboard is assembled using existing BI reports or data exploration visualizations as building blocks to assemble a summary dashboard. In many cases, a single summary dashboard can be further customized with additional filtering and analysis. Summary dashboards can be similar to a component dashboard, but a distinct difference is the use of reports and visualizations as the foundation and a smaller library of controls for application-like experiences. 

Image Source: Microsoft: PowerBI

PROS CONS
Flexibility to manage data configuration and design workflow to explore data. Simplified workflow that is better for team-based projects. Limitation for customization and typically missing components or business logic for more complex workflows and user experiences.


3. Widget dashboards

Widget dashboards provide pre-packaged analysis capabilities and/or data-source specific modules that can be assembled and deployed quickly very quickly. Visualization and connectivity widgets are properly paired and feature interactivity and visualization best practices already built in. Modern widget dashboard solutions are extendable for data acquisition and business logic.

Image Source: DataBox

PROS CONS
Widget dashboards benefit from best practices and powerful built-in features and eliminate the chance customer created defects and maximize adoption through applied best practices. Widget based dashboards can be limited in data sources, form factors, or configurability, but that is typically by design to ensure fast design and deployment.


4. KPI Tile dashboards

A new take on report / exploration dashboards are KPI tile dashboards, which provide an additional layer of summarization, and emphasizes for most important figures before drilling into supporting detailed analytics. This KPI tile dashboard workflow is increasingly popular, as more cloud data sources are required to paint a complete picture of organizational metrics. 

Image Source: Domo

PROS CONS
KPI tile dashboards are highly effective for management and executive dashboards and create fast navigation between a wide range of sources and subjects. As the number of tiles grows so does the time required to navigate, explore, and trends and locate the most important figures.


5. Embedded Analytics / Dashboards

Most cloud software solutions today include their own embedded version of “dashboards” which can range from a single time series charts to a full-blown dashboard and reporting solution. For enterprises that don’t build their own dashboard capabilities from the ground up, there are lots of options for building dashboards. Embedded dashboards provide context to application workflows and are fine-tuned for the application data / metadata. Options range from data visualization libraries to full blown embedded BI platforms.

Image Source: LogiAnalytics

PROS CONS
Embedded dashboards put analytics in the hands of users where they work, which in many cases can be integrated closer to business workflows and action. Most embedded dashboard capabilities are limited in scope and ability to leverage data from outside of the platforms for which they are embedded.


Features are Moving Targets

Another common approach for selecting the right dashboard / data viz solution is feature evaluation and comparison. It is difficult to grade vendors because the solutions are perpetually changing and improving. Additionally, many BI vendors use a platform (multiple tools) approach to cover the broad set of capabilities an enterprise may require.

Below, is a small fraction collection of technical features you could consider for your dashboard approach. 

Skills required (analyst, statistical, SQL, design, coding, scripting, CSS, etc) Desktop or cloud design and deployment Time to create and update dashboards with data readily available. Data Connectivity options Excel, web service, on-premise, cloud, proprietary
IT maintenance required Software footprint (memory, servers, cores) Scalability impact on performance and footprint Data volume restrictions client and server
Extendable: Client visualizations, data connections, logic, and platform Multi-data source support, joining, blending. Drilling extents. Can you drill to transactional data? Support and learning resources from vendor.
Community size, and engagement model Embeddable inside 3rd party apps and portals (Sharepoint, Salesforce.com, etc.) Business logic and scripting requirements End user filtering, ranking
Visual conditional formatting vs business threshold / alerts management OLAP analysis Public and private cloud options Security options and SSO. Built in security integration or API – based?
Proprietary data connectivity Big data- Spark / Hadoop Support for statistical languages like R Localization, language support
Geospatial functions and analytics and support for spatial data types Custom web services REST, SOAP, etc. Mobile phone and/or tablet support by OS native or HTML5 Data source joining, abilities.
Data preparation and blending Collaboration features, email, annotation, sharing, curation  Ability to embed in other applications Data filtering and analysis components
Sophistication of tabular data analysis (crosstab, vertical table) Search – client and/or server Included models for forecasting, trending What if analysis calculations and user inputs
Basic visualizations: Bar, pie, line, stacked, scatter plot Advanced visualizations for multi-dimensional analysis, organizational Diagrams, flowcharts, hierarchical visualizations Level of customization for charts and user experience

What’s Next?

All of these features and approaches beg the question, “what’s next?.” In my next post, I will start to address some of the outcomes and actions that we expect from dashboards and data exploration, leading us further down a path to uncover what’s next for analytics.

What’s your Current Approach?

We want to hear what approach your organization is using and how it is working for you!

Please note: I reserve the right to delete comments that are offensive or off-topic.