Health meeting v2_20120529_kp

23
SOA Health Meeting - June 2012 Business Case for Business Intelligence Kevin Pledge [email protected] 416 949 8920 @kevinpledge http://ca.linkedin.com/in/kevinpledge

description

 

Transcript of Health meeting v2_20120529_kp

Page 1: Health meeting v2_20120529_kp

SOA Health Meeting - June 2012

Business Case for

Business Intelligence

Kevin Pledge

[email protected]

416 949 8920

@kevinpledge

http://ca.linkedin.com/in/kevinpledge

Page 2: Health meeting v2_20120529_kp

Kevin Pledge

• CEO and Co-Founder of Insight Decision Solutions

• Chair of the SOA Entrepreneurial Actuaries Section

• Member of working group on Actuaries in Business Analytics

[email protected]

• 416 949 8920

• @kevinpledge

• http://ca.linkedin.com/in/kevinpledge

Page 3: Health meeting v2_20120529_kp

Agenda

Kevin: Examples and case studies of BI at work

Neil: Developing trends in semantics and big data

Page 4: Health meeting v2_20120529_kp

What is Business Intelligence?

1) Combination of data management and analytics

2) A database managed by IT

3) The ability for an organization to take all its capabilities and convert them into knowledge

4) An oxymoron

5) All of the above

Page 5: Health meeting v2_20120529_kp

Typical BI Architecture

Data Warehouse /

OLAP Server

Presentation

Server

Users

Metadata

Data Store

Integrated Systems

e.g. valuation system

ETL

ETL Source

systems

But others are possible….

Page 6: Health meeting v2_20120529_kp

‘single version of the truth’

Page 7: Health meeting v2_20120529_kp

Reality Check

data definitions, data cleansing, aggregations, normalizing, de-normalizing, rationalizing, extracts, transformations, translations,

data dictionaries, restructuring, conversions,

Not worth anything if don’t have a use for it

Page 8: Health meeting v2_20120529_kp

consistent cleansed data for actuarial

Before: Separate data extracts for experience studies, valuation, financial reporting schedules Separate rules applied for experience studies, valuation, financial reporting schedules • Date of termination in the future, policy status

terminated • Terminated policies with active coverages • Change of primary insured’s gender on joint life plans • Inconsistencies between age, DOB and issue date • System conversion errors $20m reserve for poor quality data

Page 9: Health meeting v2_20120529_kp

consistent cleansed data for actuarial

After: Single extract, single set of rules Reduction in head count, but increased number of actuaries No year-end panic Increased work capacity – more frequent reporting, more analysis Increased responsibilities, need to develop new skills Released $20m reserve for poor quality data And… I could never work in a traditional actuarial environment again

Page 10: Health meeting v2_20120529_kp

Improving Analytics

sales reporting

salesforce analysis

accounts

underwriting analysis

claim analysis

inforce demographics

valuation analysis valuation data extracts

experience studies

earnings-by-source

actu

aria

l in

volv

eme

nt

analysis frequency

retention analysis

customer analysis

Page 11: Health meeting v2_20120529_kp

Improving Analytics

Traditional Approach

BI Approach

Prepare Model / Calcs

Apply Data Report

Prepare Data Model / Calcs Analysis

Page 12: Health meeting v2_20120529_kp

Improving Analytics: Real Life Example

Medicare Supplement rate increases

$1bn premium income

~80% of business applied for increases

Assuming 6% increase

80% x $1bn x 0.06 x 1/12 = $4m

20% x $1bn x 0.06 = $12m

Page 13: Health meeting v2_20120529_kp

Improving Analytics

New frontiers

Hard to justify ROI Potentially highest ROI

Page 14: Health meeting v2_20120529_kp

Marketing Understanding your agents and customers…

Coverage

Agent

Policy

Claim

Group

Customer

Sales Team

Office

Coverage

Agent

Policy

Claim

Group

CustomerSales

Service

Page 15: Health meeting v2_20120529_kp

Underwriting – Predictive Modeling

Automate for faster, more consistent

decisions

• Reduce cost, NTU’s

• Process to improve

• over time

Page 16: Health meeting v2_20120529_kp

Claim Analysis

Meaningful attributes (ICD codes)

Alerts

Claim transition (LTC)

Testing new procedures such Wiley Protocol

Page 17: Health meeting v2_20120529_kp

Extend with Collaboration

Page 18: Health meeting v2_20120529_kp
Page 19: Health meeting v2_20120529_kp

Analytical Competitors

1. Senior executives strongly advocate analytics and fact-based decision making

2. Widespread use of descriptive statistics, predictive modeling, and complex optimization techniques

3. Analytics used across multiple business functions

4. Enterprise-wide approach to analytical tools, data, and process

Page 20: Health meeting v2_20120529_kp

Is there really a magic recipe?

Management Support

Widespread and in multiple business areas

But…

Does not need to be complex

Coordinated, not centralize control

Page 21: Health meeting v2_20120529_kp

Opportunities for Actuaries in Business Analytics

As part of the SOA's strategic objective to create and promote new areas of practice, this initiative will determine whether significant opportunities exist for actuaries in business analytics and, if so, identify projects the SOA should undertake to explore and develop these opportunities.

Session tomorrow Actuaries in Advanced Business Analytics

Page 22: Health meeting v2_20120529_kp

Summary

• Consistent cleansed data

• Improve analytics

– Accelerated

– More depth

– Lower cost

• Extend analytical decision making

Page 23: Health meeting v2_20120529_kp

Thank You

Kevin Pledge

[email protected]

416 949 8920

@kevinpledge

http://ca.linkedin.com/in/kevinpledge