3+ years
- Data are all from companies using PIMS: torture-tested by being used for a real exercise. Note only 29% of businesses analysed have data of sufficient quality to be added to the research database (3+ years of history, full data set, clear SBU definition). Data transformed into ratios and growth rates and not identified by company.
- Clustering methods used to identify matching businesses in the database with the same strategic profile (significant factors outside management control). Then split them into winners and losers by performance and look for patterns of statistically significant differences between the two groups.
Multivariate and AI tools
-
Multivariate and AI tools used to explore and quantify performance drivers, including nonlinear and interactive factors, for a variety of profitability, growth, and cost factors. Direction of causality explored by time-series tests. Collinearity eliminated by various methods. Result = models to quantify expected performance (par) plus strengths and weaknesses (positive and negative impacts of drivers on this business’s par). identified by company.
- Clustering methods used to identify matching businesses in the database with the same strategic profile (significant factors outside management control). Then split them into winners and losers by performance and look for patterns of statistically significant differences between the two groups.
