Big Data Analytics: Answering the Unanswered Questions

Transcription

1 1

2 Big Data Analytics: Answering the Unanswered Questions Session 302 IASA 86 TH ANNUAL EDUCATIONAL CONFERENCE & BUSINESS SHOW

3 Introductions John Runte, Principal Baker Tilly Virchow Krause, LLP Analytics Practice Leader Designs and Implements Analytic Solutions Many in the Insurance Industry Focuses on Big Data Solutions 3

4 Agenda 1) What is Big Data? - Characteristics - Disruptive Technologies - Where Does Big Data Fit? - Big Data At Work 2) Insurance Industry Perspectives The secret of change is to focus all of your energy, not on fighting the old, but on building the new. Socrates 4

5 What is Big Data? Big Data Is No Data Left Behind Any Data, Any Source Your Data: Decisions based on your data Transactions Documents ERP, CRM, DW/DM Big Data: Decisions based on all data relevant to you Social Data 80% Machine-Generated Data 5

6 What is Big Data? Characteristics of Big Data Velocity Volume Variety 6

7 What is Big Data? Big Data Challenges Why big data presents a technical challenge to existing BI architectures? Unfettered growth Terabytes, Petabytes, Zetabytes From many sources Analog sensors, GIS, Twitter, Social Media Relevant versus non-relevant New technical solutions different than traditional BI 7

8 What is Big Data? Answering the next unanswered question Looking for answers when you don t know the question 1) You Don t Know The Problem In Advance 2) Each Problem is Slightly Different 3) Each Problem Has Different Parts; Each With Own Problems 4) Velocity of Change & Disruption Is Significant 5) Data From All Sources» Structured Sources» Unstructured Enterprise Content» External 3rd Party Data 8

9 Disruptive Technology Application Software Over Time Differentiated ERP Technology solutions get commoditized Innovations get implemented at competitors as they learn of new developments ERP Commodity technology becomes table stakes and offers relatively low strategic value or competitive advantage Commoditized ERP Traditionally many firms focus on buying low-risk industry standard solutions Few choose to be innovators because they have the luxury to be fast followers Tactical 9 Strategic

10 Disruptive Technology The Pre-meditated Analytics Visionary Differentiated Commoditized BI BI BI BI Sees the accelerated speed of software obsolescence Possesses an accelerated innovation capability Constantly creates newer analytics to maintain distance over competitors Has made the shift to an iterative analytics world Understands that being a fast follower doesn t work Tactical 10 Strategic

11 Disruptive Technology You have a choice Pioneers Speed of Innovation Become Disrupted BI BI BI BI Become a Disrupter 11 Laggards Tactical Analytics Solution Value Strategic

12 Where Does It Fit? All Data, Unanswered Questions Search & Information Discovery OLAP Cubes Historic Source of Truth Data Warehouse / Data Marts Reporting, Query and Analysis Tools Business Intelligence Tools Advanced Analytics Prescriptive Actions, Machine Learning Data Warehouse Data Marts Sensors RSS Feeds & Social Media PAS, Claims, CRM & Other Transactional Apps Databases Unstructured Data File Systems 12 Content Mgt Systems Next Generation Databases

13 Big Data at Work.. A Critical Foundation Illustrative Example Automating an Information Discovery capability can help solve difficult problems relative to risk and appropriate pricing for risk by making sense of complicated data landscapes. Using information discovery as the foundation for your analytics will enable you to weed out the noise and return accurate risk models with more accurate pricing. To accomplish this evolution, your analytics platform should include the following: Information Discovery Foundational Capabilities Information Discovery + Other Advanced Analytics Data Discovery Efficiently expose and retrieve data whether it is structured or unstructured, internal or external data buried in the deep web or buried in a disparate company databases. Data Understanding Retrieving the data isn t enough it s the ability to understand and normalize the data that allows you to make heads from tails. Filtering and Attribute Identification Data is processed through both context and risk-specific filtering to enable insight into the most pervasive data themes. Attribute Selection What is important when? Risk discovery is not a science built on paranoia despite what traditional search engines strive for when building long lists of search results; not all data is relevant. Risk Models, Ratings Engines & Beyond.. Attributes be darned, a clear picture in the right context is what will allow your organization to think ahead of their competitors or at least more accurately price risk. INDUSTRIALIZE Communicate Measure Deliver Initiate 13 Risk & Pricing: Target Quantifiable Improvements in Combined Ratios

14 Big Data at Work.. A Critical Foundation Illustrative Data Sources Contact Management Call Center Transcripts Agent Notes Policy Generation Policy Documents Service Performance Relationship Strategy Loss Evaluation & Adjudication Compliance Management Reserve Management Satisfaction Surveys Social Network Data Adjustor Notes & Reports Loss Description Service Feedback Prior Loss Details 14

15 Unanswered Questions and Problem Solving What did I do? 1) Searched for insights 2) Got some additional information unstructured information 3) Looked at financial and transactional data 4) A combination of search, analysis, financial and unstructured data

16 AN ILLUSTRATIVE INSURANCE INDUSTRY OVERVIEW

17 The insurance industry is currently in a state of flux Shifts in growth from developed countries to emerging markets Increase in the development and consumption of data and technology Escalation of both risk and regulation Changes in customer behavior Changes in competitive landscape Big Data can help but is not a best practice, as most carriers are stymied by the challenges of making a business case Oracle Corporation Confidential Oracle Internal

18 moving to a New Normal Shifts in growth: Aging Population New Infrastructure Investments Rising Middle Class Low-Income Rural Communities Attractive Growth Markets and Segments Changes in customer behavior: Loyalty on the Decline Clear Channel Preference Heterogeneity in Emerging Markets Multi-Dimensional View of Customer a Necessity 18

19 moving to a New Normal Changing Face of the Competition: Return of the Broker Rise of the Aggregator New Players From Payer to Provider The New Normal: Slow GDP Growth Increased Regulation Commoditization Stronger Roles of Intermediaries Consumerization of IT: Going Mobile Digital Marketing Analytics 19

20 ..moving to a New Normal Includes border-line profitability, primarily via investments, and the beginning of an increasingly unfavorable interest rate environment. 20

21 Clinging to a model-driven culture vs. being Data-Driven Competitors Traditional BI Technology Stack Analytical Applications Pre-built BI Solution for Understanding Operations Operational Data Model Dashboards & Metrics Pre-built Reports Information Discovery Platform Best platform for Unstructured Analytics Server Hybrid Search/Analytical Database Flexible Data Model Structured Operational Data ERP CRM EDW Enterprise Data Warehouse 3 rd Party Content Systems, Files, Unstructured Data Social Media Big Data 21

22 Increasingly considering Big Data for competitive advantage that will yield growth and profitability Personal Lines Use of Telemetrics / Telematics, especially for discounting Emergence of Usage-based pricing M2M / Real-Time Analytics Using new insights into customer preferences and behaviors to substantially improve cross-selling and retention initiatives Analyzing 100% of the calls from the call center via audio analysis Customer-centric versus siloed by product Real-time credit and fraud scoring 22

23 Increasingly considering Big Data for competitive advantage that will yield growth and profitability Commercial Lines Lagging personal lines in innovation Use of Telemetrics / Telematics, especially for discounting Emergence of Usage-based pricing M2M / Real-Time Analytics Using new insights into customer preferences and behaviors to substantially improve cross-selling and retention initiatives Analyzing 100% of the calls from the call center via audio analysis Video for commercial property surveillance Customer-centric versus siloed by product Real-time fraud and credit scoring 23

24 Providing value across the entire insurance business process value chain 24

25 Examples Customer Acquisition How and where to prospect for good business. Effective customer acquisition efforts can help companies predict which prospects are likely to respond to a specific marketing campaign. Target Marketing Where to find the right customers, how to identify good customers. Optimal target marketing efforts are broader than customer acquisition efforts and focus on defining which prospects are likely to yield profitable business. Broker Management To help your company understand how your agency or broker force is performing, and how much of their good business is being submitted to your company 25

26 Examples Cost Reduction To reduce costs, for example, by making decisions on when to use outside data and when not to. Retention Management To understand which of your company s clients are most likely to leave and, in addition, which are likely to be profitable and unprofitable Claims Process Management To understand which claims are likely to be fraudulent, which are likely to develop into large claims, and other factors 26

27 The emergence of Usage Based Insurance (UBI) A Market Study "Insurers have been experimenting with telematics for 15 years. Now, telematics is rapidly gaining momentum, and every auto insurer should be thinking about their plans for telematics and usage-based insurance." - Mark Breading, co-author of UBI study and Strategy Meets Action (SMA) Partner 27

28 The emergence of Usage Based Insurance (UBI) A Market Study Highlights of the Strategy Meets Action study include: Seventy percent of North American property and casualty insurers are engaged in some stage of usage-based insurance, whether they are operating active programs, conducting pilots, or building strategies. Eight of the top 10 US companies now have UBI programs or pilots underway. The most frequently used variables in UBI programs are mileage, time of day, speed, and hard braking. The key challenges insurers face in building successful UBI programs are managing technology costs, understanding loss experience, creating consumer demand, and overcoming patent issues. Source: Strategy Meets Action 28

29 A typical claims process Notice of Loss Fraud Management Validation and Investigation Litigation Repudiation Invoke Reinsurance Reserve Repair Salvage Subrogation CRM Update Policy Refine Reserve Data Warehouse Settle 29

30 Providing Root Cause Analysis Sales and Marketing Why do some geographical regions buy more of your product than others? How competitive is your pricing? How do agents and consumers regard your brand vis-à-vis your competitors? Loss Risk / Claims Analysis What additional factors are associated with claim losses? What additional factors are associated with fraudulent claims? Agent and Broker Management What additional attributes are associated with high-performing agents? What do agents and brokers think about your brand? 30

31 Complementing traditional BI reporting By including root-cause analysis to support level appropriate measures across the organization to enable better decision making and measures taken Client Financial Performance Measures CFO Net Sales Results by Line of Business Results by Product Results by Agency / Agent Lines of Business Strategic Business Units Operating Units Net Sales Results by Strategic Business Unit Results by Product Results by Product/SBU Results by Agency / Agent Net Sales Results by Product Results by Agency / Agent Results by Customer/Product Results by Operating Unit Net Sales Results by Product Results by Agency / Agent Results by Customer/Product Results by Operating Unit Support Consistent Evaluation of Business Performance Management Decisions Linked to Value Drivers 31

32 Insurance Based Use Cases Presenting numerous potential opportunities to better regulate risk; identify, attract and retain profitable customers and agents; offer new products and innovative pricing; etc. Opportunity Priority Sales and Marketing Customer Acquisition and Retention 2 Fraud and Abuse 5 Agent / Broker Management 4 Underwriting / Loss Risk Analysis 3 Marketing 6 Product R&D 8 Claims Management and Resolution 7 Call Center 9 Complementary to Traditional BI 1 32

33 Sample impact - Detail Discovering New Insights In Sales and Marketing for Customer Acquisition and Retention Opportunity How Big Data can help Accessing and analyzing consumer data for sales and marketing, especially customer acquisition and retention. This opportunity is aimed at propelling growth by accessing, aggregating and analyzing all types of consumer behavioral data, as captured on the web in blogs, connections, associations, travel, profiles and click activity. This will accelerate an insurer s customer acquisition process so they can increase their overall sales and improve customer retention and profitability. Using its unique technical design, big data technology can accommodate the volumes, velocity and complexity of consumer prospect and customer data, including text, and integrate data from various sources to provide new insights in sales and marketing for customer acquisition and loyalty. Additionally, it can enable further advertising insight by leveraging blogs, call center transcripts, customer surveys, customer transaction data and notes. Moreover, it can be used to provide insight and high visibility into the online channel, a key driver of growth and competitive positioning for any insurance company. Solutions that allow product visibility and insight into all parts of the channel facilitates product development, introduction and distribution. For a $3 Billion company, a 1% increase in sales effectiveness using Big Data Technology would result in ~$30 million in new sales, or ~ $2.5 million to the bottom line with a 8.5% profit margin. 33

34 Sample impact - Detail Discovering New Insights for Underwriting and Loss / Risk Analysis Opportunity This opportunity is aimed at accessing, aggregating and analyzing all types of underwriting data to assist underwriters find relevant, accurate data points by scouring through millions of sources, both on and offline. Algorithms would be used to parse the data to find what s relevant, classify the risks based on their impact, understand and assess their likelihood so each applicant s data could be subjected to them through an underwriting model. Then the highest-ranked information would then be presented to the underwriter could make informed decisions more efficiently. How Big Data can help Using its unique technical design, big data technology can accommodate the volumes, velocity and complexity of riskrelated data, including risk correlation factors, telemetrics, prospect behavior, GIS, audio, video, etc., and integrate the data from various sources to provide new insights for underwriting. It can scour blogs, reviews, news, legal filings, and relevant information from a variety of other sources including the insurer s own data, and then deliver the summarized results in a clean, categorized profile that can be saved for future reference or used to dig deeper into each data point. For a typical $3 Billion insurer with a 70% loss ratio, a 1% decrease in losses using Big Data Technology would result in material impacts to the bottom line. 34

35 Disclosure Pursuant to the rules of professional conduct set forth in Circular 230, as promulgated by the United States Department of the Treasury, nothing contained in this communication was intended or written to be used by any taxpayer for the purpose of avoiding penalties that may be imposed on the taxpayer by the Internal Revenue Service, and it cannot be used by any taxpayer for such purpose. No one, without our express prior written permission, may use or refer to any tax advice in this communication in promoting, marketing, or recommending a partnership or other entity, investment plan, or arrangement to any other party. Baker Tilly refers to Baker Tilly Virchow Krause, LLP, an independently owned and managed member of Baker Tilly International. The information provided here is of a general nature and is not intended to address specific circumstances of any individual or entity. In specific circumstances, the services of a professional should be sought Baker Tilly Virchow Krause, LLP 35

36 Questions? John Runte, Principal Baker Tilly Virchow Krause, LLP

37 Please complete the Session Evaluation Form on the conference app and include your conference Registration ID# to be included in a drawing for a free conference registration for the 2014 Annual Conference! NOTE: Your conference Registration ID# is located at the bottom left hand corner of your badge. IASA 86 TH ANNUAL EDUCATIONAL CONFERENCE & BUSINESS SHOW