The Capability Maturity Model (CMM) of the Software Engineering Institute bas played a seminal role in the field of software process improvement (SPI). Most companies actively interested in SPI will have some familiarity with the model, and indeed CMM assessments are now legally required in the US for government software procurement. The model was, however, developed for defence contract work involving large organisations, and its generalisation to other software development contexts is

more »

... e problematic (e.g. in-house projects, small software houses). This paper concerns the experiences of a representative sample of UK companies, untypical of the traditional application domain of the CMM, who have actively considered the adoption of the CMM. All but one rejected the model. The results are analysed in terms of Roger's seminal work on technology transfer. The main reason for rejection was the lack of a prior engineering culture in the companies, compounded by the failure to achieve short-term gains. External imposition of the CMM was another inauspicious factor. It is concluded that, for the CMM to achieve wider adoption, it must be made simpler, more easily tailored and more readily linked to business goals.