1. Introduction
As technology advances and the topic of going global takes center stage in many company expansion plans, the key factors of executing a successful business plan and product marketing campaign still lies within the fundamental concepts of:
a. Trading area of a given place of doing business
b. Economies of scale in procurement and distribution cost
c. Foreseeable volume of demand
d. Product positioning and market positioning
The key differentiation between a successful business or otherwise depends largely on how robust, comprehensive and adaptable are the strategies adopted by its management team to address the above factors. The purpose of this project is to examine whether seemingly obvious assumptions based on hearsay, news, and qualitative observations about specific market conditions in Singapore still hold true when a data-driven analytical approach is taken instead to presume those assumptions.
This project aims to show aspiring local entrepreneurs, overseas franchisees, or even owners of an existing family business with a long history in Singapore how basic applied data science may possibly help them stand out from the competition by “being at the right place, selling the right thing”, through insights about existing data.
The type of business to be discussed in this project is – supplier of essential ingredients for Asian cuisines.
2. Data required and the source of the data.
The key data required for the purpose of this project are as follows:
a. Locations of selected neighborhoods in Singapore using https://www.gps-coordinates.net/ to assess the possible operational issues that come with traveling time and distance.
b. No of residents within each neighborhood to determine the foreseeable demand using data from www.singstats.gov.sg.
c. Foursquare data to identify common venues by consumers.
d. K-means clustering to segment data from Foursquare, based on similar characteristics.
With the outcomes from the above data processing, the information produced will help businesses make an informed decision and strategize on where the business should be conducted to achieve the highest potential revenue at the lowest possible cost.
3. Methodology
a. Import the necessary libraries in Pandas
b. Download and explore the data set of Singapore residents
c. Use geopy library to get the latitude and longitude values of Singapore.
d. Create a map of Singapore with 10 neighborhoods superimposed on top.
e. Define Foursquare Credentials and Version
f. Explore Neighborhoods in Singapore
g. Analyze each Neighborhood
h. Obtain top 10 most common venues of each neighborhood
i. Run k-means to cluster the neighborhood into 5 clusters.
j. Examine the 5 clusters to identify segmentation and locations that will achieve the highest potential in revenue and economies of scale for a business.
4. Results section
Based on the results, it is interesting to compare the trade-offs and advantages between Cluster 2 & 4 in terms of the distance between each similar neighborhoods and the total average no of residents over 2 years (ie demand volume).
5. Discussion, observation and recommendation
Though Cluster 2 presents the best potential for revenue justified by the high demand in volume which will be generated by the average no of residents in the neighborhoods, the distance between the neighborhoods within that same cluster is rather far. For this reason, more distribution costs may be incurred during daily operations. Moreover, the first common venue within that cluster is a food court or coffee shop. What this means is that the low market prices of food at these forms of eateries will in turn expect the suppliers of essential ingredients to sell at the corresponding low prices.
In contrast, Cluster 4 has Japanese and Chinese Restaurants as the most common venues based on Foursquare data. Based on the location map plotted, those two neighborhoods and the Central Business District are within a relatively shorter distance as compared to other neighborhoods within their respective clusters. In terms of daily operations, this means that the business in Cluster 2 stands to benefit from the economies of scale in terms of distribution and overhead costs.
6. Conclusion
In conclusion, the outcome of this project recommends that the most suitable location to conduct a business to supply essential ingredients for Japanese and Chinese Cuisines is within downtown core areas near Anson Road, Chinatown, and Fort Canning. Further to this project, other important factors like property rental price can be included as well as increasing the no. of neighborhoods in the clustering process to make the outcome more comprehensive and accurate.