An interactive pre-contract dispute avoidance toolkit for New Zealand construction industry

Purpose Avoiding disputes in the New Zealand (NZ) construction industry is crucial but challenging. While many studies focus on post-contract measures, there is limited research on pre-contract dispute avoidance, especially in traditionally procured projects, which are more prone to disputes in NZ. Therefore, the purpose of this is to study investigate pre-contract steps to avoid potential disputes in NZ and present the findings in an interactive toolkit. Design/methodology/approach An analysis of 35 construction-related court cases in NZ was conducted to identify the causes of disputes, and 14 expert interviews were carried out to collect recommendations for the pre-contract stage to avoid potential disputes. The gathered data were analysed manually, then recorded and coded using NVIVO software. Microsoft Power BI was used for data modelling and toolkit interface development (data visualisation). During one-to-one interviews and focus group sessions, 16 industry practitioner responses were considered to validate and refine the toolkit. Findings The toolkit has been located on an online platform, allowing users to access it conveniently. The finalised toolkit provides a user-centred and flexible way to select the pre-contract stage, party and theme to visualise dispute avoidance steps and their relationships towards the causes of disputes. Originality/value The toolkit offers guidance for avoiding disputes in traditionally procured projects before construction begins, focusing on contractual understanding, practices, contract formation and expectations. The interactive nature of the toolkit assists practitioners from contractor, client and consultants to configure dispute avoidance steps in a user-friendly manner. The tender preparation and pricing stage is identified as the stage where more focus is worthwhile to minimise potential post-contract disputes. This toolkit could evolve into new versions for other procurement methods, and its outcomes could train language models for improved conversation handling.