Top 10 Ways to be Successful Implementing Predictive ...€¦ · 9/16/2009 Top 10 Ways to be...
Transcript of Top 10 Ways to be Successful Implementing Predictive ...€¦ · 9/16/2009 Top 10 Ways to be...
Top 10 Ways to be Successful Implementing Predictive Modeling in Insurance Commercial Markets
October 21, 2009
Underwriting Transformation
© Z
uri
ch
In
su
ran
ce
Co
mp
an
y
2Top 10 Ways to be Successful Implementing Predictive Modeling in Insurance Commercial Markets9/16/2009
Zurich
Global footprint
– Five continents
– 170+ countries
3rd largest commercial insurer
26 straight quarters of profitability
© Z
uri
ch
In
su
ran
ce
Co
mp
an
y
3Top 10 Ways to be Successful Implementing Predictive Modeling in Insurance Commercial Markets9/16/2009
What is the Value Proposition related to Predictive Analytics?
Design risk based pricing structure
Focus retention effort on profitable risks
Identify potential premium leakage
Implement straight through processing
Optimize audit procedures, loss control, service offerings
Enhance our ability to detect fraud
Focus sales on profitable niches
Target new business
Identify highly profitable agencies
Reduce application information collection
Increase our Cross-selling / Product Density
Loss Ratio
Improvement
Expense
Reduction
Profitable Growth
Value Drivers: Implementation Methods:
Customer
Centricity
© Z
uri
ch
In
su
ran
ce
Co
mp
an
y
4Top 10 Ways to be Successful Implementing Predictive Modeling in Insurance Commercial Markets9/16/2009
Purpose
Why are we here?
– Bigger wins
– Broader capabilities
What are the takeaways?
A. Analytical Transformation
B. Business Fundamentals
C. Communicate
© Z
uri
ch
In
su
ran
ce
Co
mp
an
y
5Top 10 Ways to be Successful Implementing Predictive Modeling in Insurance Commercial Markets9/16/2009
A – Analytical Transformation
Strategic Evolution of a Predictive Analytics Practice
Plant the seeds
Condition the soil
Pull the weeds
Reap what you sow
© Z
uri
ch
In
su
ran
ce
Co
mp
an
y
6Top 10 Ways to be Successful Implementing Predictive Modeling in Insurance Commercial Markets9/16/2009
A – Analytical Transformation
Industry Specific Considerations - Commercial insurance
Market Cycles
– Right price for the risk vs. what the market will bear
– Competition
– Marketing insight
Underwriting Actions
– Directional
– Continuum of actions
– Renewal strategies
Frequency vs. Severity
– Frequency based = more stable; more credible
– Severity based = less stable; less credible
Market Forces
– Expense reduction
– Customer expectations
– Loss inflation trend
© Z
uri
ch
In
su
ran
ce
Co
mp
an
y
7Top 10 Ways to be Successful Implementing Predictive Modeling in Insurance Commercial Markets9/16/2009
Average Complete Market Cycle lasts 5-10 Years
* Rate Response Triggers
1) ISO Rate changes and/or 2) Annual pricing projections
0 1 2 3 4 5 6 7 8 9 10
Market Softening
Average soft market lasts 2 – 6 years
Market Hardening
Average hard market lasts 3 – 4 years
Typical Duration of Rate Response = 21+ Months
(17%-35% of the market cycle)
Filing Process
6 Months
Premium change
15 Months
*Trigger
N Months
A – Analytical Transformation
Commercial insurance Market Cycles
© Z
uri
ch
In
su
ran
ce
Co
mp
an
y
8Top 10 Ways to be Successful Implementing Predictive Modeling in Insurance Commercial Markets9/16/2009
A – Analytical Transformation
Expense Reduction
Current Future
Review of deductibles
Collection of Financials
Calculation of Quick ratios
Providing the Financial Ranking
Lowering costs and Increasing Premiums Audited
Non-
Value
Added
Non-
Value
Added
Putting Predictive Analytics in the Underwriting Process
© Z
uri
ch
In
su
ran
ce
Co
mp
an
y
9Top 10 Ways to be Successful Implementing Predictive Modeling in Insurance Commercial Markets9/16/2009
A – Analytical Transformation
Commercial insurance Underwriting Actions
Under-priced
Significant
Rate
Increase
Great
Potential
Credit
Moderate
Potential
Debit
Good
Renew
Risk
© Z
uri
ch
In
su
ran
ce
Co
mp
an
y
10Top 10 Ways to be Successful Implementing Predictive Modeling in Insurance Commercial Markets9/16/2009
Potential for Adverse Development
64.5%
65.0%
66.0%
67.0%
64.0%
0%
25%
50%
75%
100%
125%
Kee
p all
Lose
bes
t 5%
Lose
bes
t 15%
Lose
bes
t 25%
Var
y Ret
entio
n
Scenario Description
Decile R
ete
nti
on
*
62.0%
63.0%
64.0%
65.0%
66.0%
67.0%
68.0%
Pro
jecte
d L
oss R
ati
o
Best 30%
Middle 50%
Worst 20%
Projected LR
A – Analytical TransformationCommercial insurance - Avoiding Adverse Selection
© Z
uri
ch
In
su
ran
ce
Co
mp
an
y
11Top 10 Ways to be Successful Implementing Predictive Modeling in Insurance Commercial Markets9/16/2009
Commercial Lines Predictive Modeling, Thomas R. Bayley, Emphasis magazine
A – Analytical Transformation
Commercial insurance – Frequency vs. Severity
© Z
uri
ch
In
su
ran
ce
Co
mp
an
y
12Top 10 Ways to be Successful Implementing Predictive Modeling in Insurance Commercial Markets9/16/2009
Discover unexpected insights
– customer segments
– products
– processes and workflows
Reuse model output
– refine segmentation
– one or more models working together
A – Analytical TransformationReuse insight where you can
© Z
uri
ch
In
su
ran
ce
Co
mp
an
y
13Top 10 Ways to be Successful Implementing Predictive Modeling in Insurance Commercial Markets9/16/2009
Level 6
Level 5
Level 4
Level 3
Level 2
Level 1
En
terp
rise S
trate
gy
Peo
ple
Data
Mo
deli
ng
To
ols
Bu
sin
ess P
rocesses
Resu
lts M
easu
rem
en
t
Predictive Product Cross Sell Capabilities
Straight Through Processing Enabled by Analytics
Reuse of Data and Output Tools for Other Models
Alignment with Rate Monitoring
Creation of an Analytics Strategic Roadmap
Tactical Solution Development
Value Generation
runs in tandem with
Capabilities Maturity
which runs in tandem with
Building Block Development
Bu
ildin
g B
locks
Ca
pa
bili
tie
sB – Business Fundamentals
Capabilities Maturity
© Z
uri
ch
In
su
ran
ce
Co
mp
an
y
14Top 10 Ways to be Successful Implementing Predictive Modeling in Insurance Commercial Markets9/16/2009
To achieve the goal of developing sustainable competitive advantage through profitable growth, the model must be fully
integrated across the organization.
Each piece of the puzzle must be enabled in order for the full benefits to be achieved.
Internal and External Communication
Roles and Responsibilities
Performance Management
Training People
Applications
Architecture
Interfaces
Infrastructure
Technology Product Management
Underwriting Processes
Customer Services
Marketing & Distribution
Data Management
Processes
Change Management
Geography Alignment & Structure
Agency Integration
Organization
B – Business Fundamentals
Implementation Overview
© Z
uri
ch
In
su
ran
ce
Co
mp
an
y
15Top 10 Ways to be Successful Implementing Predictive Modeling in Insurance Commercial Markets9/16/2009
B – Business Fundamentals
Data
© Z
uri
ch
In
su
ran
ce
Co
mp
an
y
16Top 10 Ways to be Successful Implementing Predictive Modeling in Insurance Commercial Markets9/16/2009
B – Business Fundamentals
Data
Cleansing / reconciling bad data
elements is time consuming
Multiple data sources required
Commercial lines
– Regulation
– Legacy systems
Data Stewardship – savvy leadership on
what the data means
© Z
uri
ch
In
su
ran
ce
Co
mp
an
y
17Top 10 Ways to be Successful Implementing Predictive Modeling in Insurance Commercial Markets9/16/2009
B – Business Fundamentals
IT Impact
Predictive tools bump
into legacy systems
which can be costly
© Z
uri
ch
In
su
ran
ce
Co
mp
an
y
18Top 10 Ways to be Successful Implementing Predictive Modeling in Insurance Commercial Markets9/16/2009
B – Business Fundamentals
Scope Control
+/- 300% +/- 100% +/- 40% +/- 10%
ConceptData gathering
Build model Build & Deploy
Start Finish
+/- 10%
+/- 10%
+/- 10%+/- 10%
© Z
uri
ch
In
su
ran
ce
Co
mp
an
y
19Top 10 Ways to be Successful Implementing Predictive Modeling in Insurance Commercial Markets9/16/2009
B – Business Fundamentals
The Last Mile
―…the most important activity for the leadership team is to keep analytical initiatives on track and to
monitor outcomes to ensure that anticipated benefits are achieved.‖Competing on Analytics, Thomas Davenport, Harvard School of Business, 2007.
© Z
uri
ch
In
su
ran
ce
Co
mp
an
y
20Top 10 Ways to be Successful Implementing Predictive Modeling in Insurance Commercial Markets9/16/2009
Enhancements
Tools
Models
Target
Segment
B – Business FundamentalsLeverage the Capabilities
$$
$ $Value
© Z
uri
ch
In
su
ran
ce
Co
mp
an
y
21Top 10 Ways to be Successful Implementing Predictive Modeling in Insurance Commercial Markets9/16/2009
B – Business Fundamentals
Gauging Success
Keep it simple - track and monitor team outcome and model success
© Z
uri
ch
In
su
ran
ce
Co
mp
an
y
22Top 10 Ways to be Successful Implementing Predictive Modeling in Insurance Commercial Markets9/16/2009
Renewal Price Change & Loss Ratio (Retail)
5.20%4.10%3.90%
6.10%4.50%4.60%
9.20%
7.30%6.40%
30.08%
29.80%30.50%30.00% 30.44%29.90%
35.30% 32.30%
35.20%
0.00%
5.00%
10.00%
15.00%
20.00%
25.00%
30.00%
35.00%
40.00%
Q1 Q2 Q3
Month
Pric
e C
han
ge &
LR
Best 3 RPC
Avg 5 RPC
Worst 2 RPC
Best 3 LR
Avg 5 LR
Worst 2 LR
B – Business Fundamentals
Gauging Success
Line of Business Countrywide—
Pricing, Loss ratio by Decile Bands
© Z
uri
ch
In
su
ran
ce
Co
mp
an
y
23Top 10 Ways to be Successful Implementing Predictive Modeling in Insurance Commercial Markets9/16/2009
Line of Business Countrywide—Retention Results by Decile Bands
Retention (Retail)
76.00% 75.10%77.00%
74.50%71.80% 70.85%
65.60%63.40% 62.10%
0.00%
10.00%
20.00%
30.00%
40.00%
50.00%
60.00%
70.00%
80.00%
90.00%
Q1 Q2 Q3
Month
Rete
nti
on
Perc
en
tag
e
Best 3
Avg 5
Worst 2
B – Business Fundamentals
Gauging Success
© Z
uri
ch
In
su
ran
ce
Co
mp
an
y
24Top 10 Ways to be Successful Implementing Predictive Modeling in Insurance Commercial Markets9/16/2009
B – Business Fundamentals
Gauging Success
© Z
uri
ch
In
su
ran
ce
Co
mp
an
y
25Top 10 Ways to be Successful Implementing Predictive Modeling in Insurance Commercial Markets9/16/2009
4. Buy-in
3. Collaboration
2. Cooperation
1. Awareness
C – Communication
Hierarchy of objectives
© Z
uri
ch
In
su
ran
ce
Co
mp
an
y
26Top 10 Ways to be Successful Implementing Predictive Modeling in Insurance Commercial Markets9/16/2009
C – Communication
It’s a People Thing
Do we have a
common
Understanding?
(Are you sure?)
© Z
uri
ch
In
su
ran
ce
Co
mp
an
y
27Top 10 Ways to be Successful Implementing Predictive Modeling in Insurance Commercial Markets9/16/2009
Implementation – Standard project stuff…
Stakeholder Analysis
Communication Plan
Status Reports
Operational Use of Analytics – A little different…
What went into this score?
What does it mean?
How do I use it?
C – Communication
Secret Sauce
© Z
uri
ch
In
su
ran
ce
Co
mp
an
y
28Top 10 Ways to be Successful Implementing Predictive Modeling in Insurance Commercial Markets9/16/2009
C – Communication
Change Management
© Z
uri
ch
In
su
ran
ce
Co
mp
an
y
29Top 10 Ways to be Successful Implementing Predictive Modeling in Insurance Commercial Markets9/16/2009
C – Communication
Complete the feedback loop – measure model effectiveness
© Z
uri
ch
In
su
ran
ce
Co
mp
an
y
30Top 10 Ways to be Successful Implementing Predictive Modeling in Insurance Commercial Markets9/16/2009
C – Communication
Share Success
© Z
uri
ch
In
su
ran
ce
Co
mp
an
y
31Top 10 Ways to be Successful Implementing Predictive Modeling in Insurance Commercial Markets9/16/2009
Top 10 List
A – Analytical Transformation1. Use Innovation and entrepreneurial principles
2. Be thought leaders
3. Reuse insight where you can
B – Business Fundamentals4. Apply project management competencies
5. Control scope – successful bite size chunks
6. Sprint the last mile – implementation is the only thing that counts
7. Measure outcomes against key performance indicators
C - Communicate8. Establish a communication plan upfront
9. Apply change management competencies to overcome adoption challenges
10. Share success
Thank you
zurich.com