ARRESTING 10 MILLION+ IMPROPER SERVER CALLS FOR COST SAVE AND ACCURATE REPORTING FOR HONG LEONG BANK

Case Summary

Goal

Rectify improper implementation for cost effectiveness and accurate reporting

People Deployed

Business Analyst, Functional Consultant, Web Analysts

Processes Deployed

Solution Design Process, Report Configuration Review Process, Variable Review Process, Event Review Process, Quality Assurance Process

Analytics Used

Adobe Analytics, Adobe DTM

Background

Hong Leong Bank Berhad is a regional financial services company based in Malaysia, with presence in Singapore, Hong Kong, Vietnam, Cambodia and China. The Bank is technology-focused and emphasises the development of financial capabilities to serve its clients across the five geographies. Hong Leong Bank Berhad (“HLB” or “the Bank”) is listed on Bursa Malaysia and forms part of the Hong Leong Group. Headquartered in Kuala Lumpur, the Bank has a strong Malaysian entrepreneurship heritage.

Challenge

To gain insights on visitor behaviour on its portal, Hong Leong Bank had procured Adobe Analytics and had implemented the same with the help of an integration partner. However, to the bank’s shock, cost of monthly server calls were 10X more than the anticipated cost. When checked with Adobe, the billing was found to be legitimate and the bank had to pay for all the server calls made. This resulted in significant loss of money and difficulty in achieving ROI. Xerago was tasked to overcome this challenge.

Solution

As part of Xerago’s Data and Digital Analytics Services, Xerago began the assignment with an assessment of Adobe’s bill to Hong Leong Bank. Assessment revealed 350,000+ exit links and 25000+ custom links were made on an average every day in the previous month resulting in billing spike.

Xerago identified the need for evaluating the implemented code at multiple levels.

Assessment of tracking code for cost efficiency revealed the following:

  • Improper usage of exit links with no mapping of internal and external domains
  • Improper usage of custom links without any conditions
  • Rules in DTM were not properly appended with conditions

All of the above resulted in manifold increase in server calls. For instance to move to a landing page from home page should ideally make only 3 server calls. Whereas the existing implementation resulted in 7 server calls.

 

Assessment of reporting efficiency revealed the following.

  • Data from the analytics platform was manually extracted and used for reporting resulting in scope for human errors
  • Overall summary dashboard was not available to understand variables and events and

These resulted in making insights generation from current reports near impossible.

Assessment of method of implementation revealed DTM was used to implement. Even though there were inefficiencies in the way implementation was done, usage of DTM made rectifications easier.

Assessment of variables assorted revealed the following:

  • Key variables like Page URL, New vs. Returning visitors etc. were missing.

This indicated existing variable implementation was similar to Google Analytics resulting in inefficient capturing of user interactions with the website

Assessment of Events Handling revealed the following:

  • Existing implementation incorporated only 7 events against 1000+ events that come out of the box with the analytics platform

This indicated only basic reporting was possible with current implementation method.

To overcome the above, custom links and exit links were optimized by deploying appropriate conditions and mapping of internal and external domains respectively. Xerago also implemented best practices and optimized the utilization of prop 3 and prop 9 traffic variables and event 7 against the traffic variables. Also, optimization of exit links leading to internal domains also prevented the inflated metrics presented in reports.

 

Xerago standardized reporting templates and automated recurring reports. Xerago also configured over 700 events to facilitate custom reporting.

Results

  • The average number of monthly exit links was brought down to 39,981 from 9,668,429
  • The average number of monthly custom links was brought down to 20,186 from 610,816
  • Hong Leong’s stakeholders got access to better and accurate reports without any inflated numbers

This might be a good time to check other Data and Digital Analytics case studies.