Not an easy question to answer, is it. Subjective terms like “better insights” or “improving decision making” don’t really cut it as they cannot be measured
Ideally you would want the dashboard that you have developed to increase revenue/decrease costs/reduce risk. But once again it is hard to quantify such metrics from dashboard usage.
Traditionally ROI is defined as
Total Profit/ Total Investment
So if you bought $100 in stocks and sold it for $150, your ROI is 50%. So if you have made a profit you have
positive ROI else you have a negative ROI
IMO the ROI for a dashboard should be defined as
Time spent on developing an automated dashboard / Time spent on developing it manually
If you had a dashboard which took 60 hours to develop in Excel and it requires 2 hrs of manual effort per week the total time spent annually on the dashboard is 60 + 56*2 = 172 hours
However if you develop a new dashboard while automating everything you might say spend 120 hours to develop it
So your ROI = 120/172 which is ~0.7 which is less than 1 and hence its a positive ROI
Before designing a new dashboard evaluate if the ROI is less than 1. These days it should be a resounding YES!