Data Analytics

Services

Examples of Practice


CostvProfit.jpeg

Profitability Analysis

A retail company came to us and asked to explore their sales data to identify meaningful trends. After we constructed a multidimensional statistical model, we were able to isolate the effect of the cost parameter on their profitability. In short we found that they were doing better selling product that required a lower cost outlay, but were doing considerably worse on higher margin goods. While trends in the data can be too noisy to see with the naked eye, statistical technology makes it possible to specify the critical aspects of the profit function and pivot towards a company’s strength.


Educational Attainment.001.png

Litigation Visualizations

During trial, arguments were presented to a jury regarding what damages should be assessed, given the education the victim might have otherwise obtained in his life. Confused by counsel’s presentation, we went online to focus test alternative presentations of the data. Randomized online participants indicated that (on average) respondents would award $300,000 less once presented with graphical representation of the relative levels of educational attainment (as measured by the U.S. Department of Labor).

Our value added here goes beyond just offering charts. We understand both the psychology of numeracy as well as experimental means to effect randomized A-B testing of visual materials for juries or bench trials. We evaluate which data will have the most impact on the jury, consider how to represent that data, and test possibilities so that clients can have maximum efficacy in (and out of) court.


Information exploration

Knowing what questions to ask is the first critical step in any project. You know your business and your cases. What we bring to the table is a rich history in knowing the right questions to ask when working through data.

The chart on the left tells this story. A firm came asked us to explain how their profitability turned on user ratings. We were able to use advanced data science to elicit a convincing analysis that products with moderate ratings were actually most profitable. “Cheap 3s” were how we pitched it—you’ve got to target those products in the middle.

It’s not enough to be able to provide the analysis. We provide a communication strategy to convey the meaning of the insights in terms that your team, client, or jury will understand.