Best Automation Project : ICICI Bank : Service2Sales

On an average ICICI Bank receives 3.6 mn calls in a month in the inbound call

center from 1.4 mn customers across Liabilities, Credit Cards and Assets base.

Of this, around 80% of the calls are for seeking information on

transactions/products. Thus there was an opportunity for the bank to use this

point of contact for varied activities such as cross-sell & upsell of other

products, retention/winback programs, obtaining information for profile

building, contact updation & linking of various existing relationships, channel

migration programs, and seeking customer referrals (Customer Get Customer

Program). This was made possible by employing a robust technology architecture

powered by high end analytics which enabled quick pitch & closure on phone by

service executive (PBO).

Amit Sethi Joint General Manager, Technology Ashish Singhal, DGM, Business Intelligence Unit QPlease highlight the key business and/or social benefits that this project has delivered. With a conservative response rate of 3% it is projected to yield 30 Mn USD in the first year itself and with learning's going forward this is expected to have a quantum leap. QWhat were the key business challenges that were faced while implementing this project? Business challenges included the following: Creation of the customer product propensity models. Training a large service workforce on selling varied products. Design of processes so as to minimize cost and offer improved convenience to customers. QWhat according to you sets this project apart from any other in its class? What's the unique selling proposition (USP) of this project? This is the first of its kind implementation in India where the inbound call centre is being used for above activities using a complex set of data analytics and technology for enhancing customer value and lifetime with the organization.

The analytics framework works by identifying the timing and content of pitch

and the technology framework helps in generation of this information and in

aiding the delivery of the same to the PBO. The core data systems herein refer

to product host systems viz. Finacle, Finone, and Prime and the central cross

sell systems Loan on Phone. The financial and non-financial transactions

information comes through these channels and the third party data (including

Credit Bureau) is used for calculating the customer's size of wallet which helps

in determining the profile of the customer.

The analytical applications use this data to arrive at rank order of the

pitch to be made to the customer and feed this information to the CRM, Finacle-CRM,

through a special application called as 'Message Centre' which is used by PBO at

the inbound call centre for knowing what to pitch to a particular customer and

also for recording the response of the customer. This information pull is done

through the Campaign Management Tool, UNICA Affinium, which is interfaced with

the CRM system. Based on the customer response, the resultant action is

reflected in either core data systems or is passed onto the field or outbound

telecalling channels for further action. For this purpose, the CRM system is

integrated with the Sales Force Automation (SFA) system which is used for lead

management. In addition to the above, on an online basis the customers who have

responded as 'Not Interested' are suppressed from the CRM screen and the next

pitch for that customer is shown to the PBO. For customers who have responded as

'Remind Me Later' this information is again stored in Data Warehouse and after

30 days a follow up call is automatically initiated to them thrugh the Campaign

Management Tool. This seamless flow of information between the core systems,

enterprise data warehouse, CRM, analytical applications, campaign management

tool and the lead management system have helped them create a robust

infrastructure which in turn has led to smooth functioning of this

project.Managing the huge technology systems integration into a data warehouse

as a 'unified comprehensive solution' for this project encountered range of

hurdles of different kinds - technical, people, change management and smooth

scalability. In term of technical hurdles, challenges existed in data accuracy,

completeness and cleansing to facilitate effective usage. Hurdles also posed in

the subsequent build up of specific logics, derived variables, data cross system

data marts and views. In term of people challenges, the project called for a

complete revisit of the 'querying philosophy'- which had to be altered to be in

line with the data design aspects as per the Sybase IQ architecture and the

domain functional requirements.

This project is expected to bring Rs 135 Cr in first year even if response

rate is as low as 3%. These expected results are amazing given total expenditure

of Rs 5.85 Cr on this project.