Purpose – Modelling users’ interactions online is envisaged to allow developers to increase the usability of online systems and will aid system developers in building better systems to meet users’ needs, hereby creating better system design processes. Design/methodology/approach – The normative task model that was developed in this paper was created through an expert review of 14 online grocery stores, using a reverse engineering technique to model the features of the stores’ ordering process. Findings – The research identified three main areas of user experience when undertaking the process of adding a product to an online trolley: attempting to retrieve the product, receiving the results of the retrieval attempt, and adding the product to the trolley. These three classifications were used as the basis for an analysis of errors. Practical implications – The findings present a model that can be used to further understand the processes of customers as they engage in an online grocery shopping visit. The normative task model presented is expected to help in the future design of online grocery stores by identifying the possible errors that users can encounter, and methods to reduce the occurrence of these errors. Errors are one area that traditional task-modeling processes ignore, due to their focus on successful processes. Originality/value – This paper presents the innovative process of the development of a normative task model for modelling user interactions when using online grocery stores.