Case Study

Key Transportation Service Brand in Indonesia Accelerates Growth with RFM Analysis and Location-Based Analytics

The Challenge

The client is a transportation service leader in Indonesia and has been serving millions of passengers in various cities in Indonesia.

As concern over the business impact of the COVID-19 pandemic escalated, the client turned to ADA to leverage our advanced data analytics and AI capabilities. The primary interest of the client was to explore XACT, ADA’s proprietary data management platform, which contains robust array of device ID based on their Identifier for Advertisers (IFA) as well as telco data. This brings forth clarity on the mobility and profiles of the commuters in their various point of interest, as well as validating several assumptions on their passenger segments.

Through comprehensive analysis on route planning and geo-strategic placement of their transportation services, the client aims to acquire deep understanding of their customer behaviour to achieve greater efficiency to accelerate growth and increase market share across all verticals.

The Strategy

ADA provided comprehensive strategy to validate several of the client’s initial assumptions on passenger’s segment using various data sources from XACT, client’s passenger, and telco data. Key techniques used in order the address the problem statement include:

Recency, Frequency, and Monetary (RFM) Analysis

The RFM analysis allows client to discover customer loyalty by analysing their usage and monetary value, prior to comparing their timeline. It is an analysis based on customer value, represented in their recency of using the service, frequency of usage, and monetary spend. Analysis of consumer behaviour pre-COVID and during COVID-19 as well as a business index were carried out to provide clarity on passenger comparison between business users and non-business users.

Point of Interest (POI) Analysis

Insights into the top 600 visited points of interest (office, malls, transport hub, etc.) in Jakarta and various satellite cities. This analysis looks into a polygonal area of the POI for granular details into the footfall density of visitors. The analysis generates insights on whether the POI location is served by client’s fleet to identify potential customers, specifically business users.

Driving Route Analysis

An insight into the client driver’s route pre-pandemic and during the pandemic within the 600 POIs.

ADA’s propriety database allows client to identify the density of commuters at different areas during specific time interval of the day with the capability of identifying home and work location, as well as POI overlay to visualise only target passengers in the Mobility Analytics Dashboard.

The Execution

ADA designed a study to analyse consumer RFM metrics across 12 months to break down the revenue distribution and contribution of users within the observed audience group as well as to access the difference pre- and post-COVID pandemic. Some of the metrics finalised include:

  1. Number of estimated users
  2. Number of rides
  3. Total revenue
  4. Percentage of users
  5. Percentage of revenue
  6. Revenue over number of users
  7. Revenue over number of rides

8 range of passengers frequency was tabulated against the attributed above, ranging from frequency of 1 trip to more than 1,000 trips for each passenger across individual month for a period of 12 months. The recency of each ride was used to further support our analysis.

Next, ADA selected the top 600 most-visited POIs in Jakarta and satellite cities within Indonesia. We geofenced the locations and separately segmented 3 types of taxi/shuttle stands for each POI according to the distance within 100m and more than 100m. All visitors’ information seen within the areas were extracted and categorised according to:

  1. Total (weekend + weekday) visitor count
  2. Weekday only visitor count
  3. Weekend only visitor count
  4. Percentage of weekend/weekday visitor count
  5. Segmentation of total business + non-business visitors
  6. Segmentation of only business-related visitors
  7. Segmentation of only non-business-related visitors
  8. Percentage of business/non-business-related visitor count

In the Mobility Analytics Dashboard, the density of commuters within various POI was configured alongside the selected time in the form of a filter and slider. The data was cross-referenced with taxi/bus stands and shuttle pick up points to gauge the adequacy of transport services in a particular area. The dashboard was preset to allow users to visualise changes in distribution of commuters in 15 minutes time intervals from 6:00 AM to 10:00 PM daily.

Demographic filters (age, gender, affluence, persona) were added to provide richer profiling and insights into target commuters.

The Results

Our findings indicate that the client’s passengers were more skewed towards business users and high affluence groups. Users who were patrons to the client’s services within the most recent month contributed 72% of the client’s annual revenues. This allows the client to consider looking into loyalty program for frequent riders or introduce a subscription business model to better leverage their userbase.

Client’s passengers consist of 7.8x more business users compared to their average competitors with 2x higher affluence, higher frequency, and recency in younger groups of business users. The most frequent user group utilises the client’s taxi services more than 10 times a day.

Within the confirmed office locations and various POI analysed, 13 out of the 60 office POIs listed had the client’s taxi stand within reasonable reach. This indicates significant business expansion opportunity for the client to provide their services in various locations across the other 47 POIs, as well as optimising route planning to strategically locate or relocate existing transportation services to capture more market share.