Food and Beverage

Data Analysis and Apriori Algorithm - Vietnamese Restaurant Menu Item Recommendations

Tags: #Food #Apriori #DataAnalysis #Tableau #Dashboard #Recommendation #Bundling #Algorithm #GoogleDataStudio #Vietnam

Background

As the Food and Beverage (F&B) industry starts to digitalise menu, orders, and payment, they need to be equipped with the right tools to generate insights from the data. While larger corporations may have dedicated teams to do this, smaller F&B companies do not have such capacity. As such, Analytico Asia partnered with a Vietnamese Restaurant in Singapore to give them actionable insights from their data. 

Solution

Analytico Asia extracted the data from the Point of Sales (POS) system. The objective was to provide data analysis with actionable insights to improve sales and customer retention. Each feature was analysed against other aspects of the business to generate insights. Analytico also applied Apriori Algorithm (Association Rule Mining) to generate potential bundling of menu items during specific period to provide the business with a more targeted approach.

Results and Evaluation

After comprehensive data cleaning and pre-processing using Python, we analysed the data in Tableau Dashboards and Google Data Studio for 2 user-friendly approaches which our end-users can easily interact with. Some of the insights that were drawn are:

From these insights, we created a heatmap to identify lull periods and identified 2-5pm as periods with significantly lower sales. We applied Apriori Algorithm to mine items that are normally bought together by customers. Using a combination of Support, Confidence, and Lift scores, we evaluated the following as potential bundling items:

Our final recommendations were to target the lull periods by giving promotions on these data-driven bundled items. There were namely:

Some of our findings are shared within the embedded dashboards below. Readers are also encouraged to self-explore other visualisations within it.