I got the idea to create a Tableau dashboard utilizing the Pareto Principle (also known as the 80/20 rule) after hearing a podcast host talk about productivity among groups/teams in the workplace. The podcast I was listening to articulated how at any given time 20% of the company's employees are responsible for producing 80% of the company's revenue, generally speaking. The Pareto Principle states that 80% of the outcomes are derived from 20% of the causes. With this concept in mind , I decided to apply the Pareto Principle to FinCen's SAR STATS anti-money laundering data and find out which counties in each US state are producing 80% of the money laundering activity in their state.

The Pareto bar chart you see above incorporates 2022 data taken from Financial Crimes Enforcement Network's (FinCen) SAR STATS website, which houses a large repository of Suspicious Activity Report(SAR) data which is published periodically. For more context, I created a new column in the dataset titled, 'Typology Category'  to provide more clarity on what some typologies are referencing to. For example the typology 'Against Financial Institutions(s)'  by itself is very vague but when mapped to the appropriate FinCen category, this typology is referencing a Cyber Event/Attack. This feature can be seen in the data visualization.‍

For more context when comparing counties within each state, I added 2020 Census data and the total suspicious activity count from the previous year (2021). This data can be seen when hovering over the bar chart when viewing the data visualization in a desktop browser.

The number totals on the data visualizations are not the actual Suspicious Activity Report (SAR) totals but rather the total suspicious activity count i.e. the typology count in all SARs. This is my interpretation of FinCen's guidance.

There are three states that do not show a bar graph when selected and these three states are Nevada, Rhode Island and South Dakota. The bar graph does not appear for these states because the calculation I used in Tableau displays the number of counties that are equal to or less than 80%. Clark County (Nevada), Providence County (Rhode Island) and Lincoln County (South Dakota) each comprise a percentage greater than 80% and therefore a bar graph will not be displayed. The same can be said for any combination of selections in the ‘Typology Category’ and ‘AML Typology’ filters.

Generally speaking,  a county with a high suspicious activity count such as Harris County, TX (Houston, TX) will translate to a county with a high Suspicious Activity Report count. The population of a county is one of the greatest factors in determining the number of SARs filed for a county. What I found interesting is that Dallas County, TX with a population of approximately 2.6 million has a lower suspicious activity count then Montgomery County, TX (a suburban county of Houston, TX) with a county population of approximately 600K. An even more interesting insight is when the dashboard is filtered only for the following two AML typologies of (1) Human Smuggling (2) Human Trafficking; Montgomery County, TX has about twice the suspicious activity count for these two typologies when compared to its much more populated neighbor Harris County, TX. My initial suspicion for this is that the high incident rate of illicit activity in Harris County, TX is spilling over into neighboring Montgomery County, TX.

I will state the obvious in that I learned how to construct a Pareto Chart. I watched a tutorial on YouTube to get the general idea but then ended up making my own version with a lot of tinkering. I continued to strengthen my skills in Python as all the data transformation was done in Python. For this dashboard, what I learned the most was maybe design principles. This dashboard was created in 2023, but this is my favorite dashboard design still to this day(Spring 2025). You can do a lot when only using a minimal amount of colors.