All Dashboards are bad, but some are useful #ND25

Discussing the disadvantages of Dashboard as a Service in Analytics with some potential solutions

Dashboards are a huge business in the Analytics Industry. In fact, There are a lot of Analytics startups that make good 🤑 only by selling Dashboard as a Service to their Clients (in the place where I live, Bangalore)

But there’s always a never-ending conflict between Coders and Dashboard developers (Note: These two roles don’t have to be mutually exclusive) about the disadvantages and advantages of developers. In this edition of Nulldata newsletter, we are going to bring you some interesting conversation about Data.

Dashboard became a key ingredient in a hot-selling food called “Self-Service Analytics” which itself is quite a polarizing topic but has sustained in the industry for quite some time and gave birth to a lot of product and service companies.

Dashboards are Dead

Dashboards have been the primary weapon of choice for distributing data over the last few decades, but they aren’t the end of the story. To increasingly democratize access to data we need to think again, and the answer may be closer than you think…! Read More>>>

I don’t like dashboards.

Emily G (not a newspaper) @EmilyGorcenski

I don’t like dashboards. There’s many reasons I don’t like dashboards and I’m gonna elaborate on one.

Some of the problems I have with dashboards: - they’re hard to version control - they’re hard to test - they let you hide code - that hidden code becomes de facto business logic - they are a terminator in an automation chain Read More>>>

“No more dashboards!”

Is it time to stop using dashboards when designing analytics applications and data products? They said, "yes." Read More>>>

Potential Solutions for Good Dashboarding

1) Create your dashboard in high-code like shiny, dash, streamlit or genie
2) the business logic should be in a separate package/module also made available as an API end point
3) Version control the code created for both back-end (business logic) and front-end (dashboard)
4) Use the available test frameworks for both back-end and front-end.
5) Do all of the above, deploy with a simple git push and enjoy a secure and scalable environment with ownR ;)
— Credit: David Kun, ownR

References:

1. Comments at https://www.linkedin.com/posts/amrrs_datascience-activity-6802845647956533248-LqEJ

2. Decline in Dashboards

I’m restarting this newsletter after being guilty of not being consistent for months. Please share your honest feedback, It’d help me see the reality of what you want to (if you want to) read and curate accordingly 🤗