SlideShare a Scribd company logo
1 of 35
Data Mining proposal
for Sahara Life Insurance
Lucknow
Submitted by
Accommodator Consultancy Services, Lucknow
Sep 28, 2013
Accommodator Consultancy Services Lucknow
Data Mining Definition
 Data Mining: According to the Gartner Group “it is
the process of discovering meaningful new
correlations, patterns and trends by sifting through
large amounts of data stored in repositories, using
pattern recognition technologies as well as
statistical and mathematical techniques”. Its part of
data warehousing which is defined below:
 Data Warehousing: is a central repository of
meaningful and accurate data created by integrating
data from disparate sources within a company, with
past and current data for both operational and
strategic decision making and senior management
reporting such as annual comparisons of agents
performance.
Accommodator Consultancy Services Lucknow
Proven Uses of Data Mining in
Life Insurance Industry
1. Rate Setting.
2. Acquiring new customers.
3. Attrition analysis/Retaining customers
4. Developing new product lines
5. Creating geographic exposure reports
6. Detecting fraudulent claims
7. Performing sophisticated campaign management
8. Estimating outstanding claim provision
9. Forecasting, planning and budgeting..
10. Understanding customer preferences, their payments of
premium and customer queries.
11. Performance Evaluation of Insurance Agents
Accommodator Consultancy Services Lucknow
1. Rate Setting
 Rate setting of each policy is an important problem for the
actuary. Traditionally likelihood and size of claim determine the
rate.
 Attributes of existing customers are automatically analyzed
(iteratively) to establish relation with claims made (or not
made), size of claim and amount disbursed.
 Individual attributes are analyzed in iterative combinations until
meaningful and practical relationship is derived.
 Variety of simple modeling techniques are available along with
visual display of results to arrive at meaningful relations.
 The goal is to categorize customers on basis of patterns of
risk, profitability and behavior. Each category can be easily
assigned a rate for known risk, profit and behavior.
Accommodator Consultancy Services Lucknow
Risk Assessment for Rate Setting
Accommodator Consultancy Services Lucknow
Price Optimization Possible
 We can develop specialized software that would let you figure
out which groups of customers are more likely to accept a
price increase and which are more likely to shop around for a
new policy.
 A 2013 marketplace survey done by Earnix, a global leader in
price optimization, found that 26 percent of all auto insurance
companies and 45 percent of the large insurance companies
(more than $1 billion in annual revenue) in North America
currently optimize their prices. An additional 36 percent of all
companies surveyed said they plan to do this in the near
future.
 We can also help you set appropriate rates for crucial urban
customers by using Big Data that we specialize in mining.
 Socio economic data associated with geospatial data would be
utilized for price optimization and informed rate decision.
Accommodator Consultancy Services Lucknow
Price Optimization Prerequisites
 Data Quantity and Quality of claims and customer should be
good.
 High Performance Analytics is required to process large amount
of data and evaluate complex what if scenarios. Special
software is required with nice visualizations.
 Competitive Intelligence – competitive landscape, pricing
strategies and customer buying preferences and demographics.
 Data exploration - Data needs to be thoroughly analyzed
through a variety of tools.
 Predictive Modeling– Insurers must use analytical tools to
perform what-if simulation and scenario testing to forecast future
behavior and improve the underwriting performance of the
insurance company.
Accommodator Consultancy Services Lucknow
2. Acquiring New Customers
 Data Mining is used to maximize marketing campaign’s
ROI by targeting customers with attributes indicative of
greater loyalty and better profits over the lifetime of
customer’s stay with the company. This ensures optimum
use of limited marketing budget which research shows
can be up to 15% of the total cost of insurance.
 Data mining can also be used to identify best time, best
season and best media to reach out to potential
customers. If the data is not being captured, we can help
setup the system to capture this data which can be very
useful.
Accommodator Consultancy Services Lucknow
3. Attrition Analysis/Retaining
Customers
 Customer attrition rate is high in insurance industry. It is far
more expensive to acquire a new customer than to retain
existing ones. Hence makes sense to retain customer.
 Data Mining can easily lead to factors that contribute to
customer attrition and predict customers likely to attrite so they
can be retained through targeted campaigns. Preventing
policy lapse is focus of all such studies. Neural network and
DT is more likely to yield good results.
 The graph shows classification.
»
Accommodator Consultancy Services Lucknow
4. Developing New Product Lines
 Products sought by customers keep changing with time.
Companies need to be on a constant lookout for change.
 To counter change, companies need to identify upfront
profitable customer profiles. New product offerings can
be tested against such profitable customers profiles.
 Once the usefulness of new product is established, it
should be prioritized for introduction to the market based
on profit, number of potential customers or speed of
acceptance.
 Gen Re in Germany taps into the vast pool of disability
data, to determine which occupations result in disabilities
for better risk assessment and appropriate products.
Accommodator Consultancy Services Lucknow
5. Geographic Exposure Report
 Insurance business and demographic database can be
augmented with socio geographic data aka spatial
attribute data.
 Purpose of doing this is to facilitate easy and informed
decision making for decision makers when setting rates
and identifying risks. Primarily used for determining
exposure and accordingly rate adjustments and
reinsurance needs.
 Data Mining tools provide for such visual reports that
facilitates quick and easy decision making.
Accommodator Consultancy Services Lucknow
6. Detecting Fraudulent Claims
 Data Mining facilitates fraudulent claims detection.
Possible saving from detecting fraud fully justifies the
investment required.
 One of the techniques employed is to profile existing data
and compare against old fraud data to accurately detect
likelihood of a bogus claim.
 Blue Cross Blue Shield saved an estimated $4 million in
1997 alone on account of saving from fraud detection.
Accommodator Consultancy Services Lucknow
7. Estimating Outstanding Claims
Provision
 In the event of huge exposure spread out among large
number of individual policies as opposed to same
exposure to limited firms, traditional methods penalize
the latter behavior thus forcing reinsurance which may be
counter productive.
 Data mining saves us from unreasonable fears by
understanding the claims and payouts for similar groups
in the past data and then predicting the real exposure.
 The aim of the modeler is to find the most granular
section of segment that results in a claim and use this
knowledge to reinsure such high risk cases.
Accommodator Consultancy Services Lucknow
8. Performing Sophisticated
Campaign Management
 As firms grow, customer centricity tends to lose focus
and instead product development takes center stage with
mass appeal for maximum profits.
 Data mining can help in identifying customer’s real needs
and desires and serves as foundation of future campaign
development.
 Data mining can also be applied to past campaign data to
understand how campaigns have done in the past to try
and improve campaigns.
Accommodator Consultancy Services Lucknow
9. Forecasting and Budgeting
 Time series modelling can help management with their
budgetary requirements.
 A report can be made that highlights the relationship
between demand as experienced by Sahara Life and a
specified set of explanatory variables pertaining to
general economy to assist forecasting. Such variables
are freely available and can help with accurate forecasts.
 A number of modelling techniques can help management
do the general planning related to finance, HR and
operations etc.
Accommodator Consultancy Services Lucknow
10. Understanding customer preferences,
payment of premiums & customer queries
 Companies have large database but purchase pattern is
usually hidden and can be uncovered using DM easily.
Ex. Chi Square Automated Interaction detection (CHAID)
for identification of profitable customers likely to persist,
predicting future behavior and enabling firms to make
proactive knowledge based decisions.
 Can be used to segment customers and then use these
segments judiciously for increasing business.
Accommodator Consultancy Services Lucknow
11. Evaluating performance of Agents /
Brokers
 Agents can be scored on various factors, including:
■ Early lapse experience and/or policies not taken up
■ Comparisons of disclosure rates identifying agents or
brokers that are good at encouraging policyholder
disclosure
■ Sales figures, such as volumes, policy size, etc.
Models are emerging that help Insurance companies manage
agents and provide incentives.
Accommodator Consultancy Services Lucknow
How US Life Insurers Use DM
Accommodator Consultancy Services Lucknow
1. Ideal underwriting is expensive with insistence on blood and urine reports for
setting price. DM can identify people at low risk who don’t need such tests.
Alternatively high risk customers can be identified who need extensive tests.
2. Determine attributes of competitor’s customers.
3. Speed up, streamline and standardize underwriting process.
4. Use third party data in conjunction with traditional underwriting for accurate
predictions. They buy data from pharmacies about prescriptions.
5. Weed out bad/unprofitable customers from good ones and find out when is a
customer about to leave.
6. Use data mining to recruit better underwriters with suitable traits by screening
their applications.
7. No legal issues are faced as it facilitates effective and efficient decisions.
8. Modeling mortality rate is impractical, hence underwriting decisions are
modeled.
9. Fraud detection.
10. Asset Liability Management.
11. Solvency Analysis.
Data Mining Process (in brief)
1. Identify Business Problem
(ex. not enough referrals availed)
2. Transform Data into Information
(collect n clean data n apply rules)
3. Take Action on Information
(design campaign for such customers)
4. Measure the Outcome
(measure campaign and remodel)
Accommodator Consultancy Services Lucknow
Data Mining Process (in detail)
Accommodator Consultancy Services Lucknow
Process: what it really means
Accommodator Consultancy Services Lucknow
 Translate business problem into one of six DM tasks.
 Locate appropriate data that can be transformed into
actionable information.
 Explore the data.
 Prepare the data by cleaning and modifying as necessary
and applying necessary rules.
 Build model, verify validity, deploy and measure results.
Demo
Accommodator Consultancy Services Lucknow
1. Campaign for Targeted Mailing – We demonstrate how to determine, from a list
of potential customers, ones most likely to buy our products, from their given
attributes and past purchasing behavior of similar customers for focused
marketing.
2. Forecasting – We demonstrates how to predict sales and other important
ratios /business indicators based on past data for better planning.
3. Market Basket analysis – We demonstrate how to determine products that are
being purchased in bundles by customers for cross selling/upselling and
controlling customer churn.
4. Sequence Analysis – We demonstrate how order of navigation on website can
be determined and how it can be leveraged for better user experience.
5. Call Center Improvement – We demonstrate how Neural Network algorithm
may be used to identify hidden patterns in previously unknown information.
.
Targeted Mailing Campaign
Accommodator Consultancy Services Lucknow
1. Attributes of existing customers are analyzed and model is trained.
2. A user specified % of records is set aside for testing at later stage.
3. Multiple algorithms are applied to same data ex: Decision Tree, Naïve Bayes,
Cluster etc.
4. Prospects likely to buy insurance along with the probability is compared across
algorithms for models validity and usefulness.
5. Lift offered by each algorithm is analyzed by comparing the models with actual
production data set aside in testing phase.
6. Ascribe a consistent holdout seed value for consistent results (due to keeping
aside records for testing at later stages).
7. A number of parameters are available for customized prediction.
8. Input columns can be continuous or discrete, though few models do not support
all ex. Naïve does not support continuous columns.
9. Prediction value based on existing customers can be easily applied to an
external table with prospective customers with similar attributes.
.
Targeted Mailing
Campaign(visual)
Accommodator Consultancy Services Lucknow
If insurance buyer has so many attributes believed to be in play
by Campaign Manager, DM algorithm determines the order of
importance of such attributes for campaigners to concentrate on.
Targeted Mailing (Decision
Tree)
Accommodator Consultancy Services Lucknow
Decision Tree rules as determined by the Algorithm.
Here complex data is split into simple tree by taking into account only top few
important attributes, rest are disregarded. Darker the node, stronger the case.
Here people with 0 cars, <44 years of age and region <> ‘North America’ are likely to buy insurance
Forecast
Accommodator Consultancy Services Lucknow
1. Time period has to be decided upfront on which the forecast will take place.
2. The time periods should conclude at same point and there should not be any
gaps. Gaps if any can be removed automatically through options in mining
framework, namely previous value, mean etc in addition to by changing source.
3. Time Series algorithm is used for forecast. It supports both short term ARTXP
and long term ARIMA as well as a blend and a host of other options for better
accuracy and customization.
4. According to TS algorithm, large fluctuations are repeated and amplified.
5. For new products or newly introduced region which don’t have enough
historical data we can average out the rest of products/regions, forecast and
apply to new dataset. Here you would need to aggregate the data to be applied
collectively to different products or regions. Target is filtered model with data for
a newly introduced table. In case of Cross Prediction use parameter
REPLACE_MODEL_CASES.
6. If new data arrives that needs to be automatically considered, use parameter
EXTEND_MODEL_CASES.
Forecast (visuals)
Accommodator Consultancy Services Lucknow
.
Market Basket Analysis
Accommodator Consultancy Services Lucknow
1. Inbuilt MS Association model does duty to aid in cross selling.
2. Support and Probability parameters are available for better control. Both are
specified in %. Support is setting the rule of minimum occurrences. Setting
probability means specifying the minimum probability for condition to be true.
Importance is calculated by engine based on usefulness of rule. Ex: setting
Support to .01% means only those cases will be returned which occur in at
least 1 out of every 100 records and remaining associations will be ignored.
3. By using Singleton prediction query, its possible to recommend an additional
product to a customer given a/set of complementary product/s he/she buys.
This recommendation comes with probability and support for better decision
making. Of course this can be automated to show recommendation for each
customer in the database in one go based on product bundles frequently
purchased.
.
Market Basket Analysis
(visuals)
Accommodator Consultancy Services Lucknow
.
Example taken from retail industry sample data !
Sequence Clustering -
Accommodator Consultancy Services Lucknow
1. Inbuilt MS Sequence Clustering model does duty to find out the sequence of
purchases in a single transaction on internet.
2. Support and Probability parameters are available for better control. Both are
specified in %. Support is setting the rule of minimum occurrences. Setting
probability means specifying the minimum probability for condition to be true.
Importance is calculated by engine based on usefulness of rule. Ex: setting
Support to .01% means only those cases will be returned which occur in at
least 1 out of every 100 records and remaining associations will be ignored.
3. By using Singleton prediction query, its possible to recommend an additional
product to a customer given a/set of complementary product/s he/she buys.
This recommendation comes with probability and support for better decision
making. Of course this can be automated to show recommendation for each
customer in the database in one go based on product bundles frequently
purchased.
.
Improving Call Center
Customer Satisfaction
Accommodator Consultancy Services Lucknow
1. We can go for Neural Network when we have no prior expectation of what data
will show (i.e. if the call center has not had any analysis done so far). We will
use this data to suggest improvements in a call center with 30 days of data
available to us. The questions that will be answered is: what factors affect
customer satisfaction and what can call centers do to improve customer
satisfaction?
2. Once we have the answers we can use logistic regression model for
predictions. It can be used to do financial scoring and predict customer
behavior based on customer demographics.
.
Improving Call Center
Customer
Satisfaction(visual)
Accommodator Consultancy Services Lucknow
Our offerings
Accommodator Consultancy Services Lucknow
1. Data Mining in Insurance – Once the business
problem/challenge has been shared with us, we
analyze the problem, identify how useful data mining
would be, and design the entire data mining solution.
2. Cloud Services – We specialize in helping our
customers move their SQL Server databases to Cloud
and suggest appropriate package based on usage and
future growth.
3. Data Cleansing and Migration Services- We specialize
in cleansing data and data quality services.
4. Text Mining – We are capable of extracting data from
social media sites and any other web sites for further
use.
5. Big Data – We can help our clients use Big Data for
decision making.
Why ACS?
Accommodator Consultancy Services Lucknow
 We have vast experience in implementing data
warehouses and data mining models in companies such
as Fidelity, Capital One, GMAC, UTC, GE, VWG and FM
Global Insurance.
 We have the skills to be able to work with Big Data
(Hadoop).
 We are based in Lucknow and will give you the attention
you deserve.
 Dedicated SME will be involved in the project along with
data mining experts.
 We believe in delivering value for money solutions and
cost would be the lowest and result based.
 Our engineers are multi faceted and can help you with
your other data related problems as well.
Questions/Comments?
Accommodator Consultancy Services Lucknow
Our contact details:
Ankur Khanna: Director Technical
945 166 8432
Dr Vibhor Mahendru: Director Business Development
800 536 5132
THANK YOU

More Related Content

What's hot

Unbundling the Insurance Value Chain - Disruption in the Insurance Sector - T...
Unbundling the Insurance Value Chain - Disruption in the Insurance Sector - T...Unbundling the Insurance Value Chain - Disruption in the Insurance Sector - T...
Unbundling the Insurance Value Chain - Disruption in the Insurance Sector - T...SigortaTatbikatcilariDernegi
 
Uses of analytics in the field of Banking
Uses of analytics in the field of BankingUses of analytics in the field of Banking
Uses of analytics in the field of BankingNiveditasri N
 
Ten Commandments for Tackling Fraud: The Role of Big Data and Predictive Anal...
Ten Commandments for Tackling Fraud: The Role of Big Data and Predictive Anal...Ten Commandments for Tackling Fraud: The Role of Big Data and Predictive Anal...
Ten Commandments for Tackling Fraud: The Role of Big Data and Predictive Anal...CA Technologies
 
Analystics in banking and financial services
Analystics in banking and financial servicesAnalystics in banking and financial services
Analystics in banking and financial servicesRoshithaSunil
 
Modernizing the Insurance Value Chain: Top Three Digital Imperatives
Modernizing the Insurance Value Chain: Top Three Digital ImperativesModernizing the Insurance Value Chain: Top Three Digital Imperatives
Modernizing the Insurance Value Chain: Top Three Digital ImperativesCognizant
 
Data driven approach to KYC
Data driven approach to KYCData driven approach to KYC
Data driven approach to KYCPankaj Baid
 
Vcu Stm Transformation 02 15 10
Vcu Stm Transformation 02 15 10Vcu Stm Transformation 02 15 10
Vcu Stm Transformation 02 15 10guesta24f4bc
 
Customer Lifecycle Engagement for Insurance Companies
Customer Lifecycle Engagement for Insurance CompaniesCustomer Lifecycle Engagement for Insurance Companies
Customer Lifecycle Engagement for Insurance Companiesedynamic
 
Vendor strategies: Operational Business Intelligence for Agile Enterprises
Vendor strategies: Operational Business Intelligence for Agile EnterprisesVendor strategies: Operational Business Intelligence for Agile Enterprises
Vendor strategies: Operational Business Intelligence for Agile EnterprisesKishore Jethanandani, MBA, MA, MPhil,
 
AI powered decision making in banks
AI powered decision making in banksAI powered decision making in banks
AI powered decision making in banksPankaj Baid
 
Expanding BIs role by including Predictive Analytics
Expanding BIs role by including Predictive AnalyticsExpanding BIs role by including Predictive Analytics
Expanding BIs role by including Predictive AnalyticsMiguel Garcia
 
2016 Cardlytics Forrester TEI Research Report
2016 Cardlytics Forrester TEI Research Report2016 Cardlytics Forrester TEI Research Report
2016 Cardlytics Forrester TEI Research ReportMatthew Christensen
 

What's hot (15)

Unbundling the Insurance Value Chain - Disruption in the Insurance Sector - T...
Unbundling the Insurance Value Chain - Disruption in the Insurance Sector - T...Unbundling the Insurance Value Chain - Disruption in the Insurance Sector - T...
Unbundling the Insurance Value Chain - Disruption in the Insurance Sector - T...
 
Uses of analytics in the field of Banking
Uses of analytics in the field of BankingUses of analytics in the field of Banking
Uses of analytics in the field of Banking
 
Ten Commandments for Tackling Fraud: The Role of Big Data and Predictive Anal...
Ten Commandments for Tackling Fraud: The Role of Big Data and Predictive Anal...Ten Commandments for Tackling Fraud: The Role of Big Data and Predictive Anal...
Ten Commandments for Tackling Fraud: The Role of Big Data and Predictive Anal...
 
Analystics in banking and financial services
Analystics in banking and financial servicesAnalystics in banking and financial services
Analystics in banking and financial services
 
BRIDGEi2i Customer Intelligence Solutions
BRIDGEi2i Customer Intelligence SolutionsBRIDGEi2i Customer Intelligence Solutions
BRIDGEi2i Customer Intelligence Solutions
 
Modernizing the Insurance Value Chain: Top Three Digital Imperatives
Modernizing the Insurance Value Chain: Top Three Digital ImperativesModernizing the Insurance Value Chain: Top Three Digital Imperatives
Modernizing the Insurance Value Chain: Top Three Digital Imperatives
 
Data driven approach to KYC
Data driven approach to KYCData driven approach to KYC
Data driven approach to KYC
 
Vcu Stm Transformation 02 15 10
Vcu Stm Transformation 02 15 10Vcu Stm Transformation 02 15 10
Vcu Stm Transformation 02 15 10
 
Customer Lifecycle Engagement for Insurance Companies
Customer Lifecycle Engagement for Insurance CompaniesCustomer Lifecycle Engagement for Insurance Companies
Customer Lifecycle Engagement for Insurance Companies
 
Vendor strategies: Operational Business Intelligence for Agile Enterprises
Vendor strategies: Operational Business Intelligence for Agile EnterprisesVendor strategies: Operational Business Intelligence for Agile Enterprises
Vendor strategies: Operational Business Intelligence for Agile Enterprises
 
AI powered decision making in banks
AI powered decision making in banksAI powered decision making in banks
AI powered decision making in banks
 
Cis 500 assignment 4
Cis 500 assignment 4Cis 500 assignment 4
Cis 500 assignment 4
 
Expanding BIs role by including Predictive Analytics
Expanding BIs role by including Predictive AnalyticsExpanding BIs role by including Predictive Analytics
Expanding BIs role by including Predictive Analytics
 
2016 Cardlytics Forrester TEI Research Report
2016 Cardlytics Forrester TEI Research Report2016 Cardlytics Forrester TEI Research Report
2016 Cardlytics Forrester TEI Research Report
 
BankSight ROI White Paper
BankSight ROI White PaperBankSight ROI White Paper
BankSight ROI White Paper
 

Similar to Data Mining Proposal for Sahara Life Insurance

Data Mining in Life Insurance Business
Data Mining in Life Insurance BusinessData Mining in Life Insurance Business
Data Mining in Life Insurance BusinessAnkur Khanna
 
Insuring the insurance business with actionable analytics
Insuring the insurance business with actionable analyticsInsuring the insurance business with actionable analytics
Insuring the insurance business with actionable analyticsWNS Global Services
 
Harnessing the data exhaust stream: Changing the way the insurance game is pl...
Harnessing the data exhaust stream: Changing the way the insurance game is pl...Harnessing the data exhaust stream: Changing the way the insurance game is pl...
Harnessing the data exhaust stream: Changing the way the insurance game is pl...Accenture Insurance
 
Harnessing the data exhaust stream: Changing the way the insurance game is pl...
Harnessing the data exhaust stream: Changing the way the insurance game is pl...Harnessing the data exhaust stream: Changing the way the insurance game is pl...
Harnessing the data exhaust stream: Changing the way the insurance game is pl...Accenture Insurance
 
On Good Behavior, Best's Review, February, 2017
On Good Behavior, Best's Review, February, 2017On Good Behavior, Best's Review, February, 2017
On Good Behavior, Best's Review, February, 2017Gates Ouimette
 
Keeping in Step With Strategic Business Objectives in Insurance through Analy...
Keeping in Step With Strategic Business Objectives in Insurance through Analy...Keeping in Step With Strategic Business Objectives in Insurance through Analy...
Keeping in Step With Strategic Business Objectives in Insurance through Analy...Vijai John
 
Gould Scholastic Award – Julian Fung, Lasse Fuss, Tommy Ng
Gould Scholastic Award – Julian Fung, Lasse Fuss, Tommy NgGould Scholastic Award – Julian Fung, Lasse Fuss, Tommy Ng
Gould Scholastic Award – Julian Fung, Lasse Fuss, Tommy NgJulian Fung
 
Data Science Use Cases in The Banking and Finance Sector
Data Science Use Cases in The Banking and Finance SectorData Science Use Cases in The Banking and Finance Sector
Data Science Use Cases in The Banking and Finance SectorSofiaCarter4
 
POV Fueling GrowthThrough Customer Centricity
POV Fueling GrowthThrough Customer CentricityPOV Fueling GrowthThrough Customer Centricity
POV Fueling GrowthThrough Customer CentricityRob Golden
 
Predictive analytics. overview of skills and opportunities
Predictive analytics. overview of skills and opportunitiesPredictive analytics. overview of skills and opportunities
Predictive analytics. overview of skills and opportunitiesFarid Gurbanov
 
Fixing the Insurance Industry: How Big Data can Transform Customer Satisfaction
Fixing the Insurance Industry: How Big Data can Transform Customer SatisfactionFixing the Insurance Industry: How Big Data can Transform Customer Satisfaction
Fixing the Insurance Industry: How Big Data can Transform Customer SatisfactionCapgemini
 
Insurance strategy: Evolving into a digital underwriter
Insurance strategy: Evolving into a digital underwriterInsurance strategy: Evolving into a digital underwriter
Insurance strategy: Evolving into a digital underwriterAccenture Insurance
 
Grow your business by using payment analytics and insights
Grow your business by using payment analytics and insightsGrow your business by using payment analytics and insights
Grow your business by using payment analytics and insightsitio Innovex Pvt Ltv
 
3+ Keys to Proactive Underwriting (1).pdf
3+ Keys to Proactive Underwriting (1).pdf3+ Keys to Proactive Underwriting (1).pdf
3+ Keys to Proactive Underwriting (1).pdfCogitate.us
 
Future-Proofing Insurance: Deepening Insights, Reinventing Processes and Resh...
Future-Proofing Insurance: Deepening Insights, Reinventing Processes and Resh...Future-Proofing Insurance: Deepening Insights, Reinventing Processes and Resh...
Future-Proofing Insurance: Deepening Insights, Reinventing Processes and Resh...Cognizant
 
Predictive analytics-white-paper
Predictive analytics-white-paperPredictive analytics-white-paper
Predictive analytics-white-paperShubhashish Biswas
 
Financial services use cases
Financial services use casesFinancial services use cases
Financial services use casesErni Susanti
 
Gmid associates services portfolio bank
Gmid associates  services portfolio bankGmid associates  services portfolio bank
Gmid associates services portfolio bankPankaj Jha
 
Telecom analytics brochure
Telecom analytics brochure Telecom analytics brochure
Telecom analytics brochure Daniel Thomas
 
Application of predictive analytics
Application of predictive analyticsApplication of predictive analytics
Application of predictive analyticsPrasad Narasimhan
 

Similar to Data Mining Proposal for Sahara Life Insurance (20)

Data Mining in Life Insurance Business
Data Mining in Life Insurance BusinessData Mining in Life Insurance Business
Data Mining in Life Insurance Business
 
Insuring the insurance business with actionable analytics
Insuring the insurance business with actionable analyticsInsuring the insurance business with actionable analytics
Insuring the insurance business with actionable analytics
 
Harnessing the data exhaust stream: Changing the way the insurance game is pl...
Harnessing the data exhaust stream: Changing the way the insurance game is pl...Harnessing the data exhaust stream: Changing the way the insurance game is pl...
Harnessing the data exhaust stream: Changing the way the insurance game is pl...
 
Harnessing the data exhaust stream: Changing the way the insurance game is pl...
Harnessing the data exhaust stream: Changing the way the insurance game is pl...Harnessing the data exhaust stream: Changing the way the insurance game is pl...
Harnessing the data exhaust stream: Changing the way the insurance game is pl...
 
On Good Behavior, Best's Review, February, 2017
On Good Behavior, Best's Review, February, 2017On Good Behavior, Best's Review, February, 2017
On Good Behavior, Best's Review, February, 2017
 
Keeping in Step With Strategic Business Objectives in Insurance through Analy...
Keeping in Step With Strategic Business Objectives in Insurance through Analy...Keeping in Step With Strategic Business Objectives in Insurance through Analy...
Keeping in Step With Strategic Business Objectives in Insurance through Analy...
 
Gould Scholastic Award – Julian Fung, Lasse Fuss, Tommy Ng
Gould Scholastic Award – Julian Fung, Lasse Fuss, Tommy NgGould Scholastic Award – Julian Fung, Lasse Fuss, Tommy Ng
Gould Scholastic Award – Julian Fung, Lasse Fuss, Tommy Ng
 
Data Science Use Cases in The Banking and Finance Sector
Data Science Use Cases in The Banking and Finance SectorData Science Use Cases in The Banking and Finance Sector
Data Science Use Cases in The Banking and Finance Sector
 
POV Fueling GrowthThrough Customer Centricity
POV Fueling GrowthThrough Customer CentricityPOV Fueling GrowthThrough Customer Centricity
POV Fueling GrowthThrough Customer Centricity
 
Predictive analytics. overview of skills and opportunities
Predictive analytics. overview of skills and opportunitiesPredictive analytics. overview of skills and opportunities
Predictive analytics. overview of skills and opportunities
 
Fixing the Insurance Industry: How Big Data can Transform Customer Satisfaction
Fixing the Insurance Industry: How Big Data can Transform Customer SatisfactionFixing the Insurance Industry: How Big Data can Transform Customer Satisfaction
Fixing the Insurance Industry: How Big Data can Transform Customer Satisfaction
 
Insurance strategy: Evolving into a digital underwriter
Insurance strategy: Evolving into a digital underwriterInsurance strategy: Evolving into a digital underwriter
Insurance strategy: Evolving into a digital underwriter
 
Grow your business by using payment analytics and insights
Grow your business by using payment analytics and insightsGrow your business by using payment analytics and insights
Grow your business by using payment analytics and insights
 
3+ Keys to Proactive Underwriting (1).pdf
3+ Keys to Proactive Underwriting (1).pdf3+ Keys to Proactive Underwriting (1).pdf
3+ Keys to Proactive Underwriting (1).pdf
 
Future-Proofing Insurance: Deepening Insights, Reinventing Processes and Resh...
Future-Proofing Insurance: Deepening Insights, Reinventing Processes and Resh...Future-Proofing Insurance: Deepening Insights, Reinventing Processes and Resh...
Future-Proofing Insurance: Deepening Insights, Reinventing Processes and Resh...
 
Predictive analytics-white-paper
Predictive analytics-white-paperPredictive analytics-white-paper
Predictive analytics-white-paper
 
Financial services use cases
Financial services use casesFinancial services use cases
Financial services use cases
 
Gmid associates services portfolio bank
Gmid associates  services portfolio bankGmid associates  services portfolio bank
Gmid associates services portfolio bank
 
Telecom analytics brochure
Telecom analytics brochure Telecom analytics brochure
Telecom analytics brochure
 
Application of predictive analytics
Application of predictive analyticsApplication of predictive analytics
Application of predictive analytics
 

Recently uploaded

The Codex of Business Writing Software for Real-World Solutions 2.pptx
The Codex of Business Writing Software for Real-World Solutions 2.pptxThe Codex of Business Writing Software for Real-World Solutions 2.pptx
The Codex of Business Writing Software for Real-World Solutions 2.pptxMalak Abu Hammad
 
08448380779 Call Girls In Diplomatic Enclave Women Seeking Men
08448380779 Call Girls In Diplomatic Enclave Women Seeking Men08448380779 Call Girls In Diplomatic Enclave Women Seeking Men
08448380779 Call Girls In Diplomatic Enclave Women Seeking MenDelhi Call girls
 
"LLMs for Python Engineers: Advanced Data Analysis and Semantic Kernel",Oleks...
"LLMs for Python Engineers: Advanced Data Analysis and Semantic Kernel",Oleks..."LLMs for Python Engineers: Advanced Data Analysis and Semantic Kernel",Oleks...
"LLMs for Python Engineers: Advanced Data Analysis and Semantic Kernel",Oleks...Fwdays
 
Unlocking the Potential of the Cloud for IBM Power Systems
Unlocking the Potential of the Cloud for IBM Power SystemsUnlocking the Potential of the Cloud for IBM Power Systems
Unlocking the Potential of the Cloud for IBM Power SystemsPrecisely
 
Transcript: New from BookNet Canada for 2024: BNC BiblioShare - Tech Forum 2024
Transcript: New from BookNet Canada for 2024: BNC BiblioShare - Tech Forum 2024Transcript: New from BookNet Canada for 2024: BNC BiblioShare - Tech Forum 2024
Transcript: New from BookNet Canada for 2024: BNC BiblioShare - Tech Forum 2024BookNet Canada
 
SQL Database Design For Developers at php[tek] 2024
SQL Database Design For Developers at php[tek] 2024SQL Database Design For Developers at php[tek] 2024
SQL Database Design For Developers at php[tek] 2024Scott Keck-Warren
 
Kotlin Multiplatform & Compose Multiplatform - Starter kit for pragmatics
Kotlin Multiplatform & Compose Multiplatform - Starter kit for pragmaticsKotlin Multiplatform & Compose Multiplatform - Starter kit for pragmatics
Kotlin Multiplatform & Compose Multiplatform - Starter kit for pragmaticscarlostorres15106
 
#StandardsGoals for 2024: What’s new for BISAC - Tech Forum 2024
#StandardsGoals for 2024: What’s new for BISAC - Tech Forum 2024#StandardsGoals for 2024: What’s new for BISAC - Tech Forum 2024
#StandardsGoals for 2024: What’s new for BISAC - Tech Forum 2024BookNet Canada
 
New from BookNet Canada for 2024: BNC BiblioShare - Tech Forum 2024
New from BookNet Canada for 2024: BNC BiblioShare - Tech Forum 2024New from BookNet Canada for 2024: BNC BiblioShare - Tech Forum 2024
New from BookNet Canada for 2024: BNC BiblioShare - Tech Forum 2024BookNet Canada
 
08448380779 Call Girls In Friends Colony Women Seeking Men
08448380779 Call Girls In Friends Colony Women Seeking Men08448380779 Call Girls In Friends Colony Women Seeking Men
08448380779 Call Girls In Friends Colony Women Seeking MenDelhi Call girls
 
Making_way_through_DLL_hollowing_inspite_of_CFG_by_Debjeet Banerjee.pptx
Making_way_through_DLL_hollowing_inspite_of_CFG_by_Debjeet Banerjee.pptxMaking_way_through_DLL_hollowing_inspite_of_CFG_by_Debjeet Banerjee.pptx
Making_way_through_DLL_hollowing_inspite_of_CFG_by_Debjeet Banerjee.pptxnull - The Open Security Community
 
Breaking the Kubernetes Kill Chain: Host Path Mount
Breaking the Kubernetes Kill Chain: Host Path MountBreaking the Kubernetes Kill Chain: Host Path Mount
Breaking the Kubernetes Kill Chain: Host Path MountPuma Security, LLC
 
Automating Business Process via MuleSoft Composer | Bangalore MuleSoft Meetup...
Automating Business Process via MuleSoft Composer | Bangalore MuleSoft Meetup...Automating Business Process via MuleSoft Composer | Bangalore MuleSoft Meetup...
Automating Business Process via MuleSoft Composer | Bangalore MuleSoft Meetup...shyamraj55
 
AI as an Interface for Commercial Buildings
AI as an Interface for Commercial BuildingsAI as an Interface for Commercial Buildings
AI as an Interface for Commercial BuildingsMemoori
 
Injustice - Developers Among Us (SciFiDevCon 2024)
Injustice - Developers Among Us (SciFiDevCon 2024)Injustice - Developers Among Us (SciFiDevCon 2024)
Injustice - Developers Among Us (SciFiDevCon 2024)Allon Mureinik
 
Understanding the Laravel MVC Architecture
Understanding the Laravel MVC ArchitectureUnderstanding the Laravel MVC Architecture
Understanding the Laravel MVC ArchitecturePixlogix Infotech
 
Pigging Solutions Piggable Sweeping Elbows
Pigging Solutions Piggable Sweeping ElbowsPigging Solutions Piggable Sweeping Elbows
Pigging Solutions Piggable Sweeping ElbowsPigging Solutions
 

Recently uploaded (20)

DMCC Future of Trade Web3 - Special Edition
DMCC Future of Trade Web3 - Special EditionDMCC Future of Trade Web3 - Special Edition
DMCC Future of Trade Web3 - Special Edition
 
The Codex of Business Writing Software for Real-World Solutions 2.pptx
The Codex of Business Writing Software for Real-World Solutions 2.pptxThe Codex of Business Writing Software for Real-World Solutions 2.pptx
The Codex of Business Writing Software for Real-World Solutions 2.pptx
 
08448380779 Call Girls In Diplomatic Enclave Women Seeking Men
08448380779 Call Girls In Diplomatic Enclave Women Seeking Men08448380779 Call Girls In Diplomatic Enclave Women Seeking Men
08448380779 Call Girls In Diplomatic Enclave Women Seeking Men
 
"LLMs for Python Engineers: Advanced Data Analysis and Semantic Kernel",Oleks...
"LLMs for Python Engineers: Advanced Data Analysis and Semantic Kernel",Oleks..."LLMs for Python Engineers: Advanced Data Analysis and Semantic Kernel",Oleks...
"LLMs for Python Engineers: Advanced Data Analysis and Semantic Kernel",Oleks...
 
Unlocking the Potential of the Cloud for IBM Power Systems
Unlocking the Potential of the Cloud for IBM Power SystemsUnlocking the Potential of the Cloud for IBM Power Systems
Unlocking the Potential of the Cloud for IBM Power Systems
 
Transcript: New from BookNet Canada for 2024: BNC BiblioShare - Tech Forum 2024
Transcript: New from BookNet Canada for 2024: BNC BiblioShare - Tech Forum 2024Transcript: New from BookNet Canada for 2024: BNC BiblioShare - Tech Forum 2024
Transcript: New from BookNet Canada for 2024: BNC BiblioShare - Tech Forum 2024
 
SQL Database Design For Developers at php[tek] 2024
SQL Database Design For Developers at php[tek] 2024SQL Database Design For Developers at php[tek] 2024
SQL Database Design For Developers at php[tek] 2024
 
Kotlin Multiplatform & Compose Multiplatform - Starter kit for pragmatics
Kotlin Multiplatform & Compose Multiplatform - Starter kit for pragmaticsKotlin Multiplatform & Compose Multiplatform - Starter kit for pragmatics
Kotlin Multiplatform & Compose Multiplatform - Starter kit for pragmatics
 
#StandardsGoals for 2024: What’s new for BISAC - Tech Forum 2024
#StandardsGoals for 2024: What’s new for BISAC - Tech Forum 2024#StandardsGoals for 2024: What’s new for BISAC - Tech Forum 2024
#StandardsGoals for 2024: What’s new for BISAC - Tech Forum 2024
 
New from BookNet Canada for 2024: BNC BiblioShare - Tech Forum 2024
New from BookNet Canada for 2024: BNC BiblioShare - Tech Forum 2024New from BookNet Canada for 2024: BNC BiblioShare - Tech Forum 2024
New from BookNet Canada for 2024: BNC BiblioShare - Tech Forum 2024
 
08448380779 Call Girls In Friends Colony Women Seeking Men
08448380779 Call Girls In Friends Colony Women Seeking Men08448380779 Call Girls In Friends Colony Women Seeking Men
08448380779 Call Girls In Friends Colony Women Seeking Men
 
Making_way_through_DLL_hollowing_inspite_of_CFG_by_Debjeet Banerjee.pptx
Making_way_through_DLL_hollowing_inspite_of_CFG_by_Debjeet Banerjee.pptxMaking_way_through_DLL_hollowing_inspite_of_CFG_by_Debjeet Banerjee.pptx
Making_way_through_DLL_hollowing_inspite_of_CFG_by_Debjeet Banerjee.pptx
 
Vulnerability_Management_GRC_by Sohang Sengupta.pptx
Vulnerability_Management_GRC_by Sohang Sengupta.pptxVulnerability_Management_GRC_by Sohang Sengupta.pptx
Vulnerability_Management_GRC_by Sohang Sengupta.pptx
 
Breaking the Kubernetes Kill Chain: Host Path Mount
Breaking the Kubernetes Kill Chain: Host Path MountBreaking the Kubernetes Kill Chain: Host Path Mount
Breaking the Kubernetes Kill Chain: Host Path Mount
 
Automating Business Process via MuleSoft Composer | Bangalore MuleSoft Meetup...
Automating Business Process via MuleSoft Composer | Bangalore MuleSoft Meetup...Automating Business Process via MuleSoft Composer | Bangalore MuleSoft Meetup...
Automating Business Process via MuleSoft Composer | Bangalore MuleSoft Meetup...
 
AI as an Interface for Commercial Buildings
AI as an Interface for Commercial BuildingsAI as an Interface for Commercial Buildings
AI as an Interface for Commercial Buildings
 
Injustice - Developers Among Us (SciFiDevCon 2024)
Injustice - Developers Among Us (SciFiDevCon 2024)Injustice - Developers Among Us (SciFiDevCon 2024)
Injustice - Developers Among Us (SciFiDevCon 2024)
 
Understanding the Laravel MVC Architecture
Understanding the Laravel MVC ArchitectureUnderstanding the Laravel MVC Architecture
Understanding the Laravel MVC Architecture
 
Pigging Solutions Piggable Sweeping Elbows
Pigging Solutions Piggable Sweeping ElbowsPigging Solutions Piggable Sweeping Elbows
Pigging Solutions Piggable Sweeping Elbows
 
E-Vehicle_Hacking_by_Parul Sharma_null_owasp.pptx
E-Vehicle_Hacking_by_Parul Sharma_null_owasp.pptxE-Vehicle_Hacking_by_Parul Sharma_null_owasp.pptx
E-Vehicle_Hacking_by_Parul Sharma_null_owasp.pptx
 

Data Mining Proposal for Sahara Life Insurance

  • 1. Data Mining proposal for Sahara Life Insurance Lucknow Submitted by Accommodator Consultancy Services, Lucknow Sep 28, 2013 Accommodator Consultancy Services Lucknow
  • 2. Data Mining Definition  Data Mining: According to the Gartner Group “it is the process of discovering meaningful new correlations, patterns and trends by sifting through large amounts of data stored in repositories, using pattern recognition technologies as well as statistical and mathematical techniques”. Its part of data warehousing which is defined below:  Data Warehousing: is a central repository of meaningful and accurate data created by integrating data from disparate sources within a company, with past and current data for both operational and strategic decision making and senior management reporting such as annual comparisons of agents performance. Accommodator Consultancy Services Lucknow
  • 3. Proven Uses of Data Mining in Life Insurance Industry 1. Rate Setting. 2. Acquiring new customers. 3. Attrition analysis/Retaining customers 4. Developing new product lines 5. Creating geographic exposure reports 6. Detecting fraudulent claims 7. Performing sophisticated campaign management 8. Estimating outstanding claim provision 9. Forecasting, planning and budgeting.. 10. Understanding customer preferences, their payments of premium and customer queries. 11. Performance Evaluation of Insurance Agents Accommodator Consultancy Services Lucknow
  • 4. 1. Rate Setting  Rate setting of each policy is an important problem for the actuary. Traditionally likelihood and size of claim determine the rate.  Attributes of existing customers are automatically analyzed (iteratively) to establish relation with claims made (or not made), size of claim and amount disbursed.  Individual attributes are analyzed in iterative combinations until meaningful and practical relationship is derived.  Variety of simple modeling techniques are available along with visual display of results to arrive at meaningful relations.  The goal is to categorize customers on basis of patterns of risk, profitability and behavior. Each category can be easily assigned a rate for known risk, profit and behavior. Accommodator Consultancy Services Lucknow
  • 5. Risk Assessment for Rate Setting Accommodator Consultancy Services Lucknow
  • 6. Price Optimization Possible  We can develop specialized software that would let you figure out which groups of customers are more likely to accept a price increase and which are more likely to shop around for a new policy.  A 2013 marketplace survey done by Earnix, a global leader in price optimization, found that 26 percent of all auto insurance companies and 45 percent of the large insurance companies (more than $1 billion in annual revenue) in North America currently optimize their prices. An additional 36 percent of all companies surveyed said they plan to do this in the near future.  We can also help you set appropriate rates for crucial urban customers by using Big Data that we specialize in mining.  Socio economic data associated with geospatial data would be utilized for price optimization and informed rate decision. Accommodator Consultancy Services Lucknow
  • 7. Price Optimization Prerequisites  Data Quantity and Quality of claims and customer should be good.  High Performance Analytics is required to process large amount of data and evaluate complex what if scenarios. Special software is required with nice visualizations.  Competitive Intelligence – competitive landscape, pricing strategies and customer buying preferences and demographics.  Data exploration - Data needs to be thoroughly analyzed through a variety of tools.  Predictive Modeling– Insurers must use analytical tools to perform what-if simulation and scenario testing to forecast future behavior and improve the underwriting performance of the insurance company. Accommodator Consultancy Services Lucknow
  • 8. 2. Acquiring New Customers  Data Mining is used to maximize marketing campaign’s ROI by targeting customers with attributes indicative of greater loyalty and better profits over the lifetime of customer’s stay with the company. This ensures optimum use of limited marketing budget which research shows can be up to 15% of the total cost of insurance.  Data mining can also be used to identify best time, best season and best media to reach out to potential customers. If the data is not being captured, we can help setup the system to capture this data which can be very useful. Accommodator Consultancy Services Lucknow
  • 9. 3. Attrition Analysis/Retaining Customers  Customer attrition rate is high in insurance industry. It is far more expensive to acquire a new customer than to retain existing ones. Hence makes sense to retain customer.  Data Mining can easily lead to factors that contribute to customer attrition and predict customers likely to attrite so they can be retained through targeted campaigns. Preventing policy lapse is focus of all such studies. Neural network and DT is more likely to yield good results.  The graph shows classification. » Accommodator Consultancy Services Lucknow
  • 10. 4. Developing New Product Lines  Products sought by customers keep changing with time. Companies need to be on a constant lookout for change.  To counter change, companies need to identify upfront profitable customer profiles. New product offerings can be tested against such profitable customers profiles.  Once the usefulness of new product is established, it should be prioritized for introduction to the market based on profit, number of potential customers or speed of acceptance.  Gen Re in Germany taps into the vast pool of disability data, to determine which occupations result in disabilities for better risk assessment and appropriate products. Accommodator Consultancy Services Lucknow
  • 11. 5. Geographic Exposure Report  Insurance business and demographic database can be augmented with socio geographic data aka spatial attribute data.  Purpose of doing this is to facilitate easy and informed decision making for decision makers when setting rates and identifying risks. Primarily used for determining exposure and accordingly rate adjustments and reinsurance needs.  Data Mining tools provide for such visual reports that facilitates quick and easy decision making. Accommodator Consultancy Services Lucknow
  • 12. 6. Detecting Fraudulent Claims  Data Mining facilitates fraudulent claims detection. Possible saving from detecting fraud fully justifies the investment required.  One of the techniques employed is to profile existing data and compare against old fraud data to accurately detect likelihood of a bogus claim.  Blue Cross Blue Shield saved an estimated $4 million in 1997 alone on account of saving from fraud detection. Accommodator Consultancy Services Lucknow
  • 13. 7. Estimating Outstanding Claims Provision  In the event of huge exposure spread out among large number of individual policies as opposed to same exposure to limited firms, traditional methods penalize the latter behavior thus forcing reinsurance which may be counter productive.  Data mining saves us from unreasonable fears by understanding the claims and payouts for similar groups in the past data and then predicting the real exposure.  The aim of the modeler is to find the most granular section of segment that results in a claim and use this knowledge to reinsure such high risk cases. Accommodator Consultancy Services Lucknow
  • 14. 8. Performing Sophisticated Campaign Management  As firms grow, customer centricity tends to lose focus and instead product development takes center stage with mass appeal for maximum profits.  Data mining can help in identifying customer’s real needs and desires and serves as foundation of future campaign development.  Data mining can also be applied to past campaign data to understand how campaigns have done in the past to try and improve campaigns. Accommodator Consultancy Services Lucknow
  • 15. 9. Forecasting and Budgeting  Time series modelling can help management with their budgetary requirements.  A report can be made that highlights the relationship between demand as experienced by Sahara Life and a specified set of explanatory variables pertaining to general economy to assist forecasting. Such variables are freely available and can help with accurate forecasts.  A number of modelling techniques can help management do the general planning related to finance, HR and operations etc. Accommodator Consultancy Services Lucknow
  • 16. 10. Understanding customer preferences, payment of premiums & customer queries  Companies have large database but purchase pattern is usually hidden and can be uncovered using DM easily. Ex. Chi Square Automated Interaction detection (CHAID) for identification of profitable customers likely to persist, predicting future behavior and enabling firms to make proactive knowledge based decisions.  Can be used to segment customers and then use these segments judiciously for increasing business. Accommodator Consultancy Services Lucknow
  • 17. 11. Evaluating performance of Agents / Brokers  Agents can be scored on various factors, including: ■ Early lapse experience and/or policies not taken up ■ Comparisons of disclosure rates identifying agents or brokers that are good at encouraging policyholder disclosure ■ Sales figures, such as volumes, policy size, etc. Models are emerging that help Insurance companies manage agents and provide incentives. Accommodator Consultancy Services Lucknow
  • 18. How US Life Insurers Use DM Accommodator Consultancy Services Lucknow 1. Ideal underwriting is expensive with insistence on blood and urine reports for setting price. DM can identify people at low risk who don’t need such tests. Alternatively high risk customers can be identified who need extensive tests. 2. Determine attributes of competitor’s customers. 3. Speed up, streamline and standardize underwriting process. 4. Use third party data in conjunction with traditional underwriting for accurate predictions. They buy data from pharmacies about prescriptions. 5. Weed out bad/unprofitable customers from good ones and find out when is a customer about to leave. 6. Use data mining to recruit better underwriters with suitable traits by screening their applications. 7. No legal issues are faced as it facilitates effective and efficient decisions. 8. Modeling mortality rate is impractical, hence underwriting decisions are modeled. 9. Fraud detection. 10. Asset Liability Management. 11. Solvency Analysis.
  • 19. Data Mining Process (in brief) 1. Identify Business Problem (ex. not enough referrals availed) 2. Transform Data into Information (collect n clean data n apply rules) 3. Take Action on Information (design campaign for such customers) 4. Measure the Outcome (measure campaign and remodel) Accommodator Consultancy Services Lucknow
  • 20. Data Mining Process (in detail) Accommodator Consultancy Services Lucknow
  • 21. Process: what it really means Accommodator Consultancy Services Lucknow  Translate business problem into one of six DM tasks.  Locate appropriate data that can be transformed into actionable information.  Explore the data.  Prepare the data by cleaning and modifying as necessary and applying necessary rules.  Build model, verify validity, deploy and measure results.
  • 22. Demo Accommodator Consultancy Services Lucknow 1. Campaign for Targeted Mailing – We demonstrate how to determine, from a list of potential customers, ones most likely to buy our products, from their given attributes and past purchasing behavior of similar customers for focused marketing. 2. Forecasting – We demonstrates how to predict sales and other important ratios /business indicators based on past data for better planning. 3. Market Basket analysis – We demonstrate how to determine products that are being purchased in bundles by customers for cross selling/upselling and controlling customer churn. 4. Sequence Analysis – We demonstrate how order of navigation on website can be determined and how it can be leveraged for better user experience. 5. Call Center Improvement – We demonstrate how Neural Network algorithm may be used to identify hidden patterns in previously unknown information. .
  • 23. Targeted Mailing Campaign Accommodator Consultancy Services Lucknow 1. Attributes of existing customers are analyzed and model is trained. 2. A user specified % of records is set aside for testing at later stage. 3. Multiple algorithms are applied to same data ex: Decision Tree, Naïve Bayes, Cluster etc. 4. Prospects likely to buy insurance along with the probability is compared across algorithms for models validity and usefulness. 5. Lift offered by each algorithm is analyzed by comparing the models with actual production data set aside in testing phase. 6. Ascribe a consistent holdout seed value for consistent results (due to keeping aside records for testing at later stages). 7. A number of parameters are available for customized prediction. 8. Input columns can be continuous or discrete, though few models do not support all ex. Naïve does not support continuous columns. 9. Prediction value based on existing customers can be easily applied to an external table with prospective customers with similar attributes. .
  • 24. Targeted Mailing Campaign(visual) Accommodator Consultancy Services Lucknow If insurance buyer has so many attributes believed to be in play by Campaign Manager, DM algorithm determines the order of importance of such attributes for campaigners to concentrate on.
  • 25. Targeted Mailing (Decision Tree) Accommodator Consultancy Services Lucknow Decision Tree rules as determined by the Algorithm. Here complex data is split into simple tree by taking into account only top few important attributes, rest are disregarded. Darker the node, stronger the case. Here people with 0 cars, <44 years of age and region <> ‘North America’ are likely to buy insurance
  • 26. Forecast Accommodator Consultancy Services Lucknow 1. Time period has to be decided upfront on which the forecast will take place. 2. The time periods should conclude at same point and there should not be any gaps. Gaps if any can be removed automatically through options in mining framework, namely previous value, mean etc in addition to by changing source. 3. Time Series algorithm is used for forecast. It supports both short term ARTXP and long term ARIMA as well as a blend and a host of other options for better accuracy and customization. 4. According to TS algorithm, large fluctuations are repeated and amplified. 5. For new products or newly introduced region which don’t have enough historical data we can average out the rest of products/regions, forecast and apply to new dataset. Here you would need to aggregate the data to be applied collectively to different products or regions. Target is filtered model with data for a newly introduced table. In case of Cross Prediction use parameter REPLACE_MODEL_CASES. 6. If new data arrives that needs to be automatically considered, use parameter EXTEND_MODEL_CASES.
  • 28. Market Basket Analysis Accommodator Consultancy Services Lucknow 1. Inbuilt MS Association model does duty to aid in cross selling. 2. Support and Probability parameters are available for better control. Both are specified in %. Support is setting the rule of minimum occurrences. Setting probability means specifying the minimum probability for condition to be true. Importance is calculated by engine based on usefulness of rule. Ex: setting Support to .01% means only those cases will be returned which occur in at least 1 out of every 100 records and remaining associations will be ignored. 3. By using Singleton prediction query, its possible to recommend an additional product to a customer given a/set of complementary product/s he/she buys. This recommendation comes with probability and support for better decision making. Of course this can be automated to show recommendation for each customer in the database in one go based on product bundles frequently purchased. .
  • 29. Market Basket Analysis (visuals) Accommodator Consultancy Services Lucknow . Example taken from retail industry sample data !
  • 30. Sequence Clustering - Accommodator Consultancy Services Lucknow 1. Inbuilt MS Sequence Clustering model does duty to find out the sequence of purchases in a single transaction on internet. 2. Support and Probability parameters are available for better control. Both are specified in %. Support is setting the rule of minimum occurrences. Setting probability means specifying the minimum probability for condition to be true. Importance is calculated by engine based on usefulness of rule. Ex: setting Support to .01% means only those cases will be returned which occur in at least 1 out of every 100 records and remaining associations will be ignored. 3. By using Singleton prediction query, its possible to recommend an additional product to a customer given a/set of complementary product/s he/she buys. This recommendation comes with probability and support for better decision making. Of course this can be automated to show recommendation for each customer in the database in one go based on product bundles frequently purchased. .
  • 31. Improving Call Center Customer Satisfaction Accommodator Consultancy Services Lucknow 1. We can go for Neural Network when we have no prior expectation of what data will show (i.e. if the call center has not had any analysis done so far). We will use this data to suggest improvements in a call center with 30 days of data available to us. The questions that will be answered is: what factors affect customer satisfaction and what can call centers do to improve customer satisfaction? 2. Once we have the answers we can use logistic regression model for predictions. It can be used to do financial scoring and predict customer behavior based on customer demographics. .
  • 33. Our offerings Accommodator Consultancy Services Lucknow 1. Data Mining in Insurance – Once the business problem/challenge has been shared with us, we analyze the problem, identify how useful data mining would be, and design the entire data mining solution. 2. Cloud Services – We specialize in helping our customers move their SQL Server databases to Cloud and suggest appropriate package based on usage and future growth. 3. Data Cleansing and Migration Services- We specialize in cleansing data and data quality services. 4. Text Mining – We are capable of extracting data from social media sites and any other web sites for further use. 5. Big Data – We can help our clients use Big Data for decision making.
  • 34. Why ACS? Accommodator Consultancy Services Lucknow  We have vast experience in implementing data warehouses and data mining models in companies such as Fidelity, Capital One, GMAC, UTC, GE, VWG and FM Global Insurance.  We have the skills to be able to work with Big Data (Hadoop).  We are based in Lucknow and will give you the attention you deserve.  Dedicated SME will be involved in the project along with data mining experts.  We believe in delivering value for money solutions and cost would be the lowest and result based.  Our engineers are multi faceted and can help you with your other data related problems as well.
  • 35. Questions/Comments? Accommodator Consultancy Services Lucknow Our contact details: Ankur Khanna: Director Technical 945 166 8432 Dr Vibhor Mahendru: Director Business Development 800 536 5132 THANK YOU