The study analysed a business big data problem in a case study and partly methodological way. In particular, the factors that influenced the order fulfilment time were investigated. From a table of nine columns and 15,388,470 rows, a time series analysis was first performed and then a Spearman correlation was used to analyse the relationship between completion time and the number of sales. The limitations of some methods (ANOVA, Random Forest, Factor Analysis of Mixed Data) were examined, and then Kruskal-Wallis and Dunn-Bonferroni tests were carried out, taking samples of sufficient and necessary size to allow for the large number of element variations in each column, and restricting the analysis to the variable values that sometimes appeared important. A number of variables in three columns had a significant effect on the evolution of the completion time and the methodology proved successful in identifying them. However, the analysis also highlighted the need to group the individual, non-metric variable values into new groups.