We used an iterative approach in this project. We began the requirements gathering process with in-person interviews and an analysis of the reports submitted by the marketers. From an investigation of how commodity prices were normalized, we were able to first determine the minimum amount of data required to calculate the net pay to ensure that data entry would be as efficient as possible. Then, we developed a prototype to validate the results and receive feedback about our proposed logic and methodology.
After receiving and including user feedback, we designed and implemented a general model capable of representing the normalization logic for each marketer. Care was taken not only to ensure the accuracy of the calculations, but also that the user interface was able to be configured to exactly match the labels of each individual marketer report for both efficiency and accuracy purposes. Moreover, a database model was developed to store as much of the data contained in the marketer reports as possible. Finally, using the spreadsheets as a basis, a report template was created so that commodity data could be pulled on demand for any specified week.
Specific efforts were put in place to ensure trust in the delivered solution. Moving from Excel to custom software brings with it the risk that the users will not trust the new software since its operation is not as transparent or familiar. Consequently, Dattivo ensured that intermediate calculations were displayed so that the origin of all data was clear, the project itself was phased in one step at a time, and supplementary reports were created to provide further visibility into the final calculations.
