Analysis of User and Merchant Dropoff for Sugar App - Background
| Background | Data Source | Methodology |
|---|
Contents
Introduction and Project Background
In this day and age of rapid modernization, local businesses have had a very hard time competing with large franchise chains such as Walmart, 7-11, Giant etc. Customers of local businesses also demand that local businesses deliver a high quality product (38%), yet offer the lowest price possible (38%) (The Consumer Barometer Survey, 2015). The lack of economies of scale, as well as the high infrastructure and marketing costs have led to the closure of many local businesses.
As such, there have been a rise of various local discount websites such as Groupon, Deal.com.sg, Lazada and Qoo10, which offer high quality products at low prices; and are frequented by many Singaporeans (42%) (The Consumer Barometer Survey, 2015). The growing trend suggests that there may be room for more growth in this area, which is where Sugar steps in.
Introduction of Sugar
Sugar is an interactive city guide that seeks to encourage a culture of exploration in Singapore and helping local small businesses get discovered.
Currently, Sugar operates in 3 countries – Hong Kong, Jakarta and Singapore. It originated from Shanghai where it has experienced tremendous success of over 250,000 users a day and hence, the founders has decided to expand its operations to Singapore in early 2014.
Sugar has discovered that product quality and price is essential in a user making a purchase decision. Thus, as a city guide with its location-based features, it hopes to revolutionise the online shopping industry, by providing high quality products, at the lowest price, at the most convenient locations for the user.
There are two main stakeholders: Merchants and Users as Sugar exists as a platform in a two-sided market. Like many other two-sided markets, it connects users and merchants and earn a premium for connecting these two groups by charging the transactions made between the two groups.
Sugar’s merchants are mainly small local businesses in Singapore. It has a large variety, including cafes, small restaurants, bars, hair salons, gyms, gift shops. The benefits for merchants is advertising to users that are in close proximity to them. As mentioned before, convenience is an important factor for a purchase decision and thus, Sugar is leveraging on this aspect.
Users derive benefits from the discounted deals on the app. For example, Sugar can offer a deal which offers a 50% discount on truffle fries at a restaurant near potential users. Since price is highlighted as an extremely important factor in a purchase decision, this can entice existing users and new users to check out deals on Sugar app. The location-based feature also provides them with ample opportunities to explore hidden gems and new establishments in their vicinity.
Differentiation from Other Apps
What differentiates Sugar from the rest of the apps are 2 features: Location-Based Recommendations and Skimming Mechanism.
Location-Based Recommendations
Sugar offers location-based advertising for local businesses while simultaneously offering attractive deals to users. As a heuristic, the app uses location information provided by the user to recommend deals. Location can be set by the user or by the in-built GPS system.
Skimming Mechanism
Another important feature of the app is the Skimming Mechanism. It allows each user to reduce the price of an item by 20 cents, hence “skimming the price”. This gives the app word-of-mouth potential as users are motivated to reduce the price further by persuading their friends to skim the price of an item. For example, if a person can get 4 other friends, the 5 of them can reduce the price of an item by $1 for everyone.
Business Problems and Motivations
As Sugar is a relatively young startup in Singapore, it has not yet attained a critical mass of user and merchant numbers. As such, user growth and user experience is vitally important. To reach this critical mass, Sugar needs to minimise user and merchant attrition, and retain vital segments of both groups in order to preserve the network effect of Sugar’s platform. This is done by enhancing the app experience for both users and merchants.
Unredeemed Vouchers
One issue Sugar faces is the issue of unredeemed vouchers. When a user applies for a voucher, there is a 7-day redemption period before it expires. Many users allow the vouchers expire. Maximising the voucher redemption rate will benefit both merchants and users, by ensuring that a faster turnaround time for products that users want, and higher profits for the merchant. This will enhance the user experience for both groups.
User Retention
To build a successful application, Sugar needs to grow as fast as possible while retaining existing users. Many users have become dormant or stop buying after the first purchase. Analysis can be applied to find out why they have turned dormant/ stopped buying and identify solutions that may be able to attract them.
Merchant Retention
Not all merchants are the same. There is a large variation between the popularity of individual merchants. Some merchants experience uptake daily whereas some have little to none at all. As a result, many of the latter may choose to drop out after a period of time. Since Sugar is in a two-sided market, a high attrition rate of merchants may result in users also ceasing usage. As Sugar has limited resources to reach out and engage merchants, high value targets can be identified to optimise the effort in reaching out to retain merchants. Further analysis can be finding out the main factors for merchant attrition.
Project Objectives
Merchants
- Identifying star merchants (performing better than expected) and laggard merchants (performing worse than expected)
- Based on Revenue and Redemption Rate
- Identifying time-series patterns (e.g. day of week and hour) and grouping merchants with similar redemption behavior together
Items
- Identifying star items (performing better than expected) and laggard items (performing worse than expected)
- Based on Impressions, Clicks and Price
- Identifying time-series patterns (e.g. day of week and hour) and grouping items with similar redemption behavior together - drill down to product level
Users
- Identifying star users (performing better than expected) and normal users using LRFM
- Length
- Recency
- Frequency
- Monetary
- Cluster users based on that
- Apply survival analysis from installation to purchase (e.g. which type of users are more likely to purchase only x times before falling off the app?)
Geospatial
- Determining the relationship between user location and merchant location (i.e. do users really go to places near them or are they more willing to travel to visit certain merchants?)
- Determining the relationship between redemption rate and merchant location
- Identifying popular areas and unpopular areas (i.e. the proportion of merchants in an area should be proportion to the number of orders it receive)
- Recommending potential locations for new merchants
- Develop a method to analyse cannibalisation rates
Scope of Project
The scope of our project includes the following:
- Data collection - Collating datasets from Sugar’s SQL database, Flurry and Localytics analytics dashboard
- Data preparation - Setting and censoring the data. We will also be filtering users and merchants based on location, where we will only focus only on the Singapore region
- Analysis of voucher redemption rates using survival analysis
- Analysis of survival curves for merchant and users
- Refinement - Get client feedback and refine the model
- Extrapolation of survival curve for forecast
- Derive recommendations and solutions for Sugar
Project Deliverables
- Project Proposal
- Mid-term Presentation
- Mid-term Report
- Final Presentation
- Final Report
- Project Poster
References
Deciding Factor in Selecting a Local Business. (n.d.). Retrieved January 10, 2016, from https://www.consumerbarometer.com/en/insights/?countryCode=SG
Benneyan JC, Lloyd RC, Plsek PE: Statistical process control as a tool for research and healthcare improvement. Qual Saf Health Care 2003, 12:458-464.
Caffrey, J., & Isaacs, H. H. (1971). Estimating the Impact of a College or University on the Local Economy.
Clarke, G. P., & Hayes, S. (2006). GIS and retail location models. Geomarketing: Methods and Strategies in Spatial Marketing, 165-186.
Daoud, R. , Amine, A. , Bouikhalene, B. , Lbibb, R. (2015). 'Customer Segmentation Model in E-commerce Using Clustering Techniques and LRFM Model: The Case of Online Stores in Morocco'. World Academy of Science, Engineering and Technology, International Science Index 104, International Journal of Computer, Electrical, Automation, Control and Information Engineering, 9(8), 1905 - 1915.
Larivière, B., & Van den Poel, D. (2004). Investigating the role of product features in preventing customer churn, by using survival analysis and choice modeling: The case of financial services. Expert Systems with Applications,27(2), 277-285.
Lu, J., & Park, O. (2003). Modeling customer lifetime value using survival analysis—an application in the telecommunications industry. Data Mining Techniques, 120-128.
Öner, Ö. (2014). Retail location.
Portela, S., & Menezes, R. Modeling Customer Churn: An Application of Duration Models.
Rochet, J. C., & Tirole, J. (2004). Two-sided markets: an overview (Vol. 258). IDEI working paper.
Schubert, S., & Lee, T. (2011). Time Series Data Mining with SAS® Enterprise Miner™ (1st ed.). SAS. Retrieved from https://support.sas.com/resources/papers/proceedings11/160-2011.pdf
Singstat.gov.sg,. (2016). Statistics Singapore - Services Survey Series 2014 - Food and Beverage Services. Retrieved 27 February 2016, from http://www.singstat.gov.sg/statistics/visualising-data/storyboards/sss-food-and-beverage-services/
The Connected Consumer Survey 2014/2015. (2015). Retrieved January 10, 2016, from https://www.consumerbarometer.com/en/graph-builder/?question=M1&filter=country:united_states,china,hong_kong_sar,korea,malaysia,singapore,australia
The Consumer Barometer Survey 2014/2015 (2015). Retrieved January 10, 2016, from https://www.consumerbarometer.com/en/insights/?countryCode=SG
Van den Poel, D., & Lariviere, B. (2003). Customer attrition analysis for financial services using proportional hazard models. European Journal of Operational Research, 157(1), 196-217.
Woodall DH: The Use of Control Charts in Health-Care and Public-Health Surveillance. J Qual Technol 2006, 38(2):89-104.
Wood, S., & Browne, S. (2007). Convenience store location planning and forecasting-a practical research agenda. International Journal of Retail & Distribution Management, 35(4), 233-255.
Zhang, G., & Chen, Y. (2007). An Integrated Data Mining and Survival Analysis Model for Customer Segmentation. In Integration and Innovation Orient to E-Society Volume 1 (pp. 88-95). Springer US.
