Design of business intelligence system for point-of-sales software in full-service restaurants

Abstract

Small and Medium-sized restaurants are faced with increasing competitions due to the rise of food reviews and food delivery mobile applications which increase accessibility and customer demands. In order to thrive in this competition, restaurant owners seek to know their performance beyond sales number as there are other aspects of restaurant management such as operations efficiency, inventory, and quality of service. However, the current offerings by various POS in the market do not provide such performance measures, making it impossible for owners to know and improve other dimensions of restaurant management. As a result, the case-study company identified the gap to fulfil this need and also an opportunity to provide a new value proposition to users, attracting new users and retaining existing ones. Thus, a business intelligence (BI) system with a dashboard was selected as the solution.

There are 3 components to build the BI system for full-service restaurants: Key Performance Indicators (KPIs), dashboard display, and data warehouse. For KPIs, a variety of performance measures of restaurants was researched from academic journals and industry experts. Critical ones were selected based on a set of criteria and grouped together according to Balanced Scorecard to form a set of Key Performance Indicators (KPIs). Next, each selected KPI was carefully enriched by context and comparison, then chosen an appropriate display medium based on data types and relationships. Then, these KPIs were grouped and arranged to fit a 1-page tablet-sized dashboard. The last component is the data warehouse which was necessary to provide a new data source that contains aggregate values and is structured appropriately for reporting and analysis on the BI system. Dimensional data model was used to design the physical data model for software developers to implement in the future.

The new design was then evaluated by 20 respondents who are the current users of the POS and current dashboard. They were asked to rate the current and the new design separately on the scale of 1 to 10. The rating results were further validated by Wilcoxon signed ranked test and it could be statistically shown that the new design is preferred by users to the current one. Users find the new dashboard more useful in providing them with relevant performance measures. Apart from the ratings, respondents also ranked their top 3 and bottom 3 KPIs and provided qualitative suggestion on the new design. Feedbacks were consolidated and the revised final version of the dashboard was created based on those feedbacks.

In summary, this research is the first to create a BI system based on POS software for full-service restaurants. The KPIs were selected to be useful and practical while the display was carefully designed based on the principle of visual perception and enriched by useful context to encourage critical analysis. The set of KPIs and the new design were well-received as they provide an effective way to equip restaurant owners and managers with valuable performance measures to help manage their restaurants better.