Research Dashboard: Empty State vs. Partial Data - What's the User-Friendly Approach?

Question

I'm currently working on redesigning a research data dashboard. I'm facing a decision on how to handle the initial data load for a specific type of data that dates back couple of years. I have two options in mind:

1. Empty State: This option would not display any visualizations initially. Instead, it would guide/inform the user with a clear call to action.

2. Partial Data: This option involves rendering data covering the past few months to a year on the initial load.

The primary objective of this dashboard is to allow users to drill down into data for a specific company while applying various filters. Currently, we load all the data when a user accesses the dashboard. However, I have concerns about this approach, as it could potentially overwhelm users with a vast amount of information.

I'm seeking advice on the best approach to ensure a user-friendly experience while still allowing users to access the data they need effectively. looking to get some insights and suggestions, greatly appreciated."

Design Concept

Answer 1

Let's look at Google Web Analytics as a case study.

Simply... provide the users with what your user research shows to be the most interesting or relevant data, then allow users to refine the output.

Providing users an "Empty State" by default, would suggest you don't have any clue as to what data is/should be useful.

Answer 2

Since you already have a current state implemented on showing all the data, the best way to make a decision is to actually do some research and collect analytics on the dashboards to help guide the decision.

Also, keep in mind that if this particular dashboard retrieves historical data, consider whether this dataset is also likely to grow in the future or if it is only confined to a specific period of time.

The argument against an empty state is that generally over time there will be some information that is accessed more frequently than others, so users don't usually want to do an extra step if there are other ways to be more efficient (e.g. be able to save particular views).

The argument against a partial state is that it might take just as many steps to get to the view that they want, or make it more difficult to get to it, depending on how you design the filters and the charts.

You are assuming that the users will have too much data to try and sort through, and that they won't get used to it overtime. It may very well be true, but you want to make sure that you are solving a valid problem so try to validate it first.