Search for a command to run...
Purpose The study aims to contribute to the training‐related literature by investigating the relationship between training and training transfer factors and company‐level training outcomes, and the relationship between the latter and company performance. Design/methodology/approach The authors propose and test a model linking constructs related to training (the volume and quality of training, supervisor support, peer support, and organisational incentives), constructs related to company‐level training outcomes (the acquisition and interpretation of information, and cognitive and behavioural changes) and company performance. In the empirical analysis the authors use structural equation modelling based on a sample of 247 service companies. Findings The study confirms a strong relationship between supervisor support and the volume and quality of training as well as between supervisor support and organisational incentives for training transfer. Organisational incentives are directly related to both studied company‐level training outcomes, i.e. the acquisition and interpretation of information, and cognitive and behavioural changes. In addition, they are also indirectly related to company performance through encouraging cognitive and behavioural changes. The volume and quality of training are related only to the acquisition and interpretation of information, while no direct relationship with company performance was found. Research limitations/implications The study sends an important message to CEOs and HR managers showing them that companies should focus not only on the volume and quality of training but also on important training‐transfer‐related factors. Originality/value The study complements previous studies dealing with the relationship between training and performance by clearly distinguishing between training and training transfer factors, and by studying their combined relationship with company‐level training outcomes and, through that, with company performance. The links in the model are addressed in a way that has not yet been used in existing literature.