As one of the leading telecommunications providers in Malaysia, Celcom Axiata Berhad is supported by its call centre and is accountable for receiving and transmitting huge amounts of requests to administer related incoming product support and enquiries from its customers. With more than 1000 call centre agents interacting with the customers around Malaysia, there are more than 80,000 calls related to product enquiries and support received every day. Therefore, to understand customer's satisfaction through the huge volumes of customers' feedback received is a challenge.

As a result, Celcom Axiata Berhad had embarked on an SMS Survey project in 2007 to resolve the issue and hence allowing them to efficiently gather customers' feedback. This was done in an effort to analyze customers' approval towards the support provided by the call centre. The objectives of the SMS Survey project were successfully achieved and had since become part of the KPI's monitoring tool to identify the key areas that could be used to improve the call centre's operation.

Problems/Challenges Faced
However in 2010, Celcom Axiata Berhad was faced with yet another challenge - call volumes received by the call centre had noticeably increased. As a result, a better sampling of data was required, which meant that more SMSes needed to be sent out. To better facilitate the requests, Celcom had identified that a flexible user interface as well as reporting and analysis tools were required as part of the SMS Survey system.

Speedminer Sdn. Bhd, was invited to perform a proof-of-concept (POC) project to determine its capability of coping with the high capacity of SMSes. The project was also performed to test its capability to perform complex analysis on the SMS Survey data as well as its ability to process them into meaningful and critical information that can be used by the company's management.

Speed SMS (One of the modules of Speedminer System) employs its core SMPP-based SMSC engine to churn out the data provided by Speedminer BI that extracts call information in a real time manner from Celcom's call centre. Each call can be monitored in a way that this information is fed simultaneously to Speed SMS as soon as the agent responsible for the call finishes interacting with a particular customer.
With this acquired information, details on the customer's mobile phone, the agent interacting with the particular customer, the call duration as well as customer interactive voice response (IVR) languages present a seamless survey mechanism. Furthermore, with Speed SMS, messages can be customized to suit the different types of customers.

Focus:

Unprecedented insight regarding customer information through Speed SMS surveys and Speed SMS

The SMSC engine in Speed SMS will then filter out the numbers to determine which ones are eligible to receive the SMS surveys. With the rule engine built in Speed SMS, calls that originate overseas or calls which do not represent an actual interaction can be screened to avoid unnecessary SMS being sent out.

Also, with multithreading technology and the lightweight SMPP protocol, Speed SMS is capable of delivering up to 15 SMSes per second by distributing its load to multiple SMSC in Celcom's network. This ensures maximum utilization and ensures that the system effectively maximizes the volumes of SMSes that can be sent out in a given period. Thus, more accurate survey results can be achieved in larger samples.

Achievements
After 3 months of Speed SMS implementation which includes the POC stage, the new enhanced system is proven to be reliable in delivering a higher volume of SMSes to the customers. In addition, survey analysis reports can now be produced in a timely manner. Due to the fact that reports provide near real-time information on the SMS survey data, not only is the management able to quickly and efficiently understand the situation that is going on in its call centre, but the management can now also decisively identify the key areas that need to be enhanced in order to improve its overall services provided by the call centre agents.

Note: In separate project, from year 2006 until now, Speedminer has been used by Celcom's call centre to analyze all the logs of incoming calls. Analysis such as waiting period per customer before the call is answered; average time spent on each call, and the number of repeat calls from the same customer as well as other performance analysis is performed.