Only in midsize companies were these two drivers (“optimization through a technically better platform” with 40 percent and “implementation of new types of use cases” with 53 percent), relatively close together. Only one in ten respondents listed costs as their main driver. This figure is surprisingly low considering that costs were identified as a reason for using Hadoop components.

Main driver for using Hadoop by company size (n=144)

BARC’s findings show that Hadoop usage is primarily driven by new analytic requirements. Costs are not viewed as a main driver. Although cost savings have been considered a decisive advantage of Hadoop for a long time, there seems to be a more differentiated view today. Hadoop can be – but is not necessarily – less expensive, depending on the use case.