Professional Documents
Culture Documents
Whitepaper: Insurance Business Analytics
Whitepaper: Insurance Business Analytics
www.hexaware.com
Whitepaper
Insurance Business Analytics
Table of Contents
1 Introduction
1.1 Current scenario in Insurance Business Intelligence
2 Abstract
2.1 What Insurance Business Analytics can do?
3 Proposed Solution
3.1 Problem Statement
3.2 Introduction to Solution
3.3 Key offerings
3.4 Tangible ROI - Business Benefits
4 Future Direction
4.1 Long-Term Focus
4.2 Conclusion
03
03
04
04
06
06
06
07
08
08
08
08
www.hexaware.com
Whitepaper
Insurance Business Analytics
1 Introduction
1.1 Current scenario in Insurance Business Intelligence
With improvement in data warehousing techniques and advancement in storage capacity more and more data is stored in an enterprise warehouse.
With improving data mining technology it is feasible to make knowledge discovery in a manner which was impossible earlier. New tool capabilities,
new business models and new data sources are constantly emerging. They help in knowledge discovery and finding actionable intelligence from
internal data.
Insurance is a data intensive industry and insurance companies create huge data. Insurance companies realize that they have huge wealth of data
which can be used to find answers to business related questions like
Finding answers to questions like these was impossible and impractical earlier because of non-availability of reliable data and technology to support
data mining. Through data mining and reporting technology it has now become possible to take a look at the enterprise data from different
dimension and derive actionable intelligence out of it.
Business Analytics is helps to align all the data within an organization to make sure data is accurate, timely, in context and available to all who need
it. There is no dearth of such software and the choice depends on utility, price and performance. Once Insurance companies understand the power
oftheir own data the requirement of knowledge discovery extends from reporting and dash-boarding to scoring and predicting.
Integration
Interpretation &
Evaluation
Industry Data - Various Source
Data Mining
Understanding
RAW DATA
Transformation
Target Data
Selection &
Cleaning
Data
Warehouse
www.hexaware.com
Whitepaper
The HexSOA Test Model (HSTM)
2 Abstract
It goes without saying that the availability of correct data at right time is the key to taking timely decisions. On top of it if the business data is
represented in most innovative manner, it helps understanding it quickly and drawing meaningful inference. Further if the business user is given
option to drill down to the granular level, they may not ask for more.
Insurance Analytics solutions are designed in a way that it can draw required data from source systems (often disparate and even excel files) on a
regular basis. For achieving this one has to innovatively design the data model which can store all necessary data required to produce reports /
dashboards. Once this data-model for insurance analytics is properly designed the key to its success lies in developing link with the data sources to
get correct transaction data into the data warehouse at the defined interval. One has to design links in a way that it gets all relevant data from different
data sources at the same time.
Other key requirement is designing the dash-boards so that it presents the data in most innovative way. The graphical representation of data allows
end user to understand business performance at a glance for which otherwise they would have to shuffle through endless reports. Further these
dashboards must allow end user to drill down to the minutest level and also download relevant data in desired format for further use. The interface
designing is done by understanding the key performance indicators (KPI) which is most relevant to the insurance companies. This can be done either
by asking the LOB managers to define the KPIs which are most relevant to them or by bringing in industry knowledge and suggesting standard KPIs.
Some examples:
Business By Channel
Time Run:7/25/2011 2:38:56 PM
Premium Analytics
750
Cost
600
Persistency Analysis
450
300
150
0
Agency
Broker
Corporate
Agent
Direct
40
32
Claims Analytics
24
16
8
0
Direct
Corporate Agent
40
40
60
80
20
Broker
100
40
60
80
20
100
60
80
20
100
www.hexaware.com
Whitepaper
Insurance Business Analytics
3 Proposed Solution
3.1 Problem Statement
In many Insurance Companies a divide exists between IT and a LOB managers need for knowledge and information. Storage systems help to
accumulate transaction data every day and IT managers are required to create data dump for business users for their analysis and reporting. Data
dump extracted from IT systems is passed on to business users which is used for analysis and reporting on periodic basis. The process is repetitive,
time consuming and prone to human errors and having one view of business data is always a challenge.
Not that the insurance companies are not mining their own data but the way they do it is inefficient and time consuming. Problem lies in the fact that
most of these data is generated by disparate systems and there is a possibility of the data present in different formats. Combining the data derived
from disparate system to gain actionable knowledge through use of spread sheets cannot deliver information showing it from different dimension. It
is also possible that after doing an intensive exercise one may not derive information which allows view to minutest level and present them in most
innovative manner.
Data model Data model supporting data marts with report-friendly three-tier hierarchy summary, highly summarized, and KPIs
ETL Extract, transform, and load technology for data movement, aggregation, scheduling, and error handling
Parameterized reporting frameworks Provide a broad range of business segmentation, dependency analysis, trending, meeting needs
such as channel management, underwriting efficiency, claims management, regulatory compliance, customer service, and distributor
management
Installation and implementation Includes installation of the software, basic customization, documentation, and application training.
Lapse Ratio
Time Run:7/25/2011 2:14:45 PM
Persistency Ratio
Time Run:7/25/2011 2:14:45 PM
40
40
40
60
40
60
80
80
20
100
Total Business
Renewed Business
Industry standard- Lapse Ratio
Lapse Ratio
60
42
15
30
Total Business
Renewed Business
Industry standard-Persistency Ratio
Persistency Ratio
New
Business
Target Premium(mn)
Achieved Premium(mn)
80
60
80
20
120
100
3.06
1.26
100
Loss Ratio(%)
Expense Ratio
Combined Ratio(%)
Modify
Policy
Servicing
40
60
20
100
100
Modify
Modify
Channel
Analytics
60
42
85
70
Select View
80
20
100
40
60
80
60
20
20
Ratio Analysis
Time Run:7/25/2011 2:14:45 PM
Target Status
Time Run:7/25/2011 2:14:45 PM
Premium
Analytics
51.18
26.78
77.96
60
30
0
April
July
June
March
May
Claims
Analytics
Insurance Analytics
Enterprise DWH
Policy Admin
System
Channel
Management
CRM
Claims Data
www.hexaware.com
Whitepaper
Insurance Business Analytics
Data Model
A_EDUCATION
A_MARITAL_STATUS
A_PROFESSION
A_GENDER
A_CLAMANT
Channel /
Marketing
A_AGE_DRIVER
A_VEHICLE_MAKE
A_VEHICLE_COLOR
A_PROPERTY_CLASS
A_PROPERTY_STATE
REPUDIATION_D
TIME_D
AGE_D
New
Business
Premium
Analytics
POLICY_TYPE_D
CLAIM_CAUSE_D
POLICY_PERIOD
ENGINE_CAPACITY
A_TIMES_D
CLAIMS_F
SEGMENT_D
Insurance
Business
Intelligence
Management
Dashboards
CLAIMS_TAT_F
CHANNEL_D
PRODUCT_D
Policy
Servicing
Claims
CLAIM_TYPE_D
A_VEHICLE_AGE
TAT_PROCESS_D
Premium Analytics
Channel Analytics
Policy Servicing Analytics
New Business Analytics
Claims Analytics
Going with the pre-defined data model helps faster implementation by 30% compared to effort required in creating specific data model
Hexaware Insurance Business Analytics effectively puts to end the dilemma of buy v/s build decision. It is based on industry standard data model
and use best of breed reporting tools.
www.hexaware.com
Whitepaper
Insurance Business Analytics
4 Future Direction
4.1 Long-Term Focus
It is not always possible to do analysis of data in real time as the aggregation of data into a data warehouse has its own challenges. Mostly today the
dashboards are available for past data, be it a day old or week depending upon the frequency on which the data is extracted into the DWH. However
it appears that the business users wouldn't want to wait for information and BI users are beginning to demand real time or near real time, which is
more in demand for data analysis relating to their business, particularly in front line operations. It is not too distant when insurance companies will
demand real time business information. Monthly and even weekly analysis will not suffice.
Other developing area is analytics offering on mobile devices. At present the BI dashboards and reports are being accessed using mobile devices
either through URL or dedicated apps. As per industry estimates roughly 30 -35% of BI will happen through mobile devices by 2014. Moreover the
trend may change from accessing reports using mobile devices to BI application developed for mobile devices.
4.2 Conclusion
Insurance is a data-rich industry; every insurance company has pile of data running into tera-bytes. But unfortunately, most of that data is either
underutilized or unutilized.
There is growing awareness and demand to extract actionable knowledge from internal data which is key to gaining a competitive advantage by
analyzing this data and getting a greater insight into their business. With help of Business Intelligence techniques Insurance firms can unlock the
intelligence contained in their operational applications - like policy administration, claims management, SPM and CRM solutions - through modern
data mining technology.
Slice and dice of operational data can throw useful insights into monitoring the performance of producer, tracing process efficiency of operations and
decipher patterns form claims data. Not only this, a deep understanding of past data can be helpful to design new products and related marketing
strategies, enabling the insurance firm to transform their own data into a wealth of information and actionable intelligence.
www.hexaware.com