9/28/2011 Analysis of Internet Purchasing Behavior October 3, 2011 CAS I F CAS In Focus Seminar S i Kevin Levitt Roosevelt Mosley, FCAS, MAAA Nick Kucera Senior Vice President Principal Consultant comScore, Inc. Pinnacle Actuarial Resources Pinnacle Actuarial Resources Experience the Pinnacle Difference! Discussion Topics � Current State of Insurance Marketing � Customer response analyses � comScore background and data � comScore background and data � Description of research � Characteristics of shoppers � Model development � Analysis of quotes submitted � Analysis of policies purchased Analysis of policies purchased � Impact of price on conversion � Unstructured data � Additional Research 2 1
9/28/2011 Current State of Insurance Marketing � Explosion in the investment that insurers are making in marketing � Many different advertising mediums are being used � Traditional, internet, social media, etc. � Has led to the understanding that customer behavior is about more than price � Customer service, reputation, convenience � Different elements are important to different market segments 3 Customer Response Analyses Marketing Effort Quoting Analysis A l i Quote Conversion Analysis Sale Lapse / Cancelation Analysis Analysis Renewal Quote Retention Analysis Renewal 4 2
9/28/2011 Customer Response Analyses � Quoting analysis: analysis of the likelihood of a prospective insured obtaining an insurance quote from you you � Conversion analysis: analysis of the likelihood of a prospective insured that has received a quote purchasing insurance from you � Lapse/Cancelation analysis: likelihood of an insured not lapsing or canceling the policy mid-term � Retention analysis: analysis of the likelihood of a current R i l i l i f h lik lih d f insured renewing with you 5 Customer Response Modeling – Challenges for Insurance Companies � Model structure and parameterization � New territory learning curve � New territory – learning curve � Priority � Internal data availability � Internal data applicability � Availability of price change data � Availability of price change data � Measuring market competitiveness � Applications 6 3
9/28/2011 comScore Background & Data 7 comScore is a Global Leader in Measuring the Digital World NASDAQ SCOR Clients 1700+ worldwide Employees 900+ Headquarters Reston, VA 170+ countries under measurement; Global Coverage 43 markets reported Local Presence 32 locations in 23 countries � V0910 8 4
9/28/2011 What We Do…. � We provide digital marketing intelligence that helps our customers make better-informed business decisions and implement more effective digital business strategies p g g � We measure the continuous online activity of 1 million people in the U.S. who have granted us explicit permission to confidentially measure their Internet usage patterns. � Our consumer panel is a representative cross-section of the U.S. population, worldwide regions and individual countries � We also have permission to: � Survey panelists � Survey panelists � Match to third-party databases � Append offline data 9 The Trusted Source for Digital Intelligence Across Vertical Markets 9 out of the top 10 9 out of the top 10 INVESTMENT BANKS AUTO INSURERS 9 out of the top 10 4 out of the top 4 INTERNET SERVICE WIRELESS CARRIERS PROVIDERS 47 out of the top 50 9 out of the top 10 ONLINE PROPERTIES PHARMACEUTICAL COMPANIES 9 out of the top 10 9 0 45 out of the top 50 45 out of the top 50 p CONSUMER FINANCE ADVERTISING AGENCIES COMPANIES 9 out of the top 10 9 out of the top 10 MAJOR MEDIA COMPANIES CPG COMPANIES V0910 10 5
9/28/2011 Auto Insurance Quote Detail � Data is captured from what panelists see using scraping technology scraping technology 11 Auto Insurance Quote Detail � Data for 5 Top Auto Insurance Company Sites – Quote – Drivers • ZIP code • Age • Bodily injury liability limits Bodily injury liability limits • Gender • Gender • Coverage package • Marital Status • Premium quoted • Industry/Occupation • Final Purchased Premium – Vehicles • Company name • Vehicle year/make/model/type • Homeownership • Vehicle use • Whether SSN entered • Annual mileage • Primary driver education • Comprehensive deductibles • Current Insurance Information • Collision deductibles • Whether currently insured – Incidents • Length of gap in coverage • Incident Type • Length of current carrier • Incident Description • Length continuously insured • Prior BI Limit 12 6
9/28/2011 Auto Insurance Quote Detail Data – Source of Traffic Data � Source of traffic is broken out into the following categories: � Paid search P id h � Natural search � Webmail sites � Other referred � Non-referred � For search, we know the search engine, click type (paid/natural), and the search phrase � For webmail sites and other referred, we know the referring site 13 Pinnacle & comScore Research Marketing Effort Population Target Description Of those that visit the website, people Auto Quote with what characteristics are more likely Insurance Initiated to start the quote process? Website Quote Visitor Quote Quote Of those that initiate the quote process, what is the likelihood that they complete Initiated Submitted it? Of those that actually submit the quote Of those that actually submit the quote, Q Quote t Policy Bind P li Bi d Sale how many complete the purchase? Submitted 14 7
9/28/2011 Characteristics of Shoppers 15 Data Summary – Means of Entry Model - Quote Submitted Means of Entry 80.0% 600 559 570 546 546 F S 70.0% N 500 r u u e b m 463 q m 60.0% b u i e 411 400 e t r 50.0% n t 46.8% c e o y d f 40.0% 300 o p 32.6% V f e i 30.0% r s s 200 Q i u 1 t 20.0% o , o t 0 r 100 9.7% e 0 s 10.0% s 0 5.4% 5.5% 0.0% 0 Natural Search Non-Referred Other Referred Sponsored Search Webmail Means of Entry Exposure Percentage Frequency per 1000 16 8
9/28/2011 Data Summary – Session Day/Time Model: Purchase Made During Session 30.0% 120.00 ure (Sessions with completed quote) 25.0% 100.00 Purchase Frequency (per 1,000) 20.0% 80.00 15.0% 60.00 10.0% 40.00 Expos 5.0% 20.00 0.0% - Weekday Day (9-5) Weekday Morning Weekday Evening Weekday Late Weekend Night Saturday Sunday (10PM+) Session Day/Time Data Summary – Driver1 Age Model: Purchase Made During Session 18.0% 100.00 90.00 re (Sessions with completed quote) 16.0% 80.00 Purchase Frequency (per 1,000) 14.0% 70.00 12.0% 60.00 10.0% 50.00 8.0% 40.00 6.0% 30.00 4.0% 4.0% Exposur 20 00 20.00 P 2.0% 10.00 0.0% - <=18 19 20 21 22 23 24 25-29 30-34 35-39 40-44 45-49 50-54 55-59 60-64 65+ Age Driver 1 18 9
9/28/2011 Data Summary – Number Vehicles/Drivers Model: Purchase Made During Session 70.0% 90.00 re (Sessions with completed quote) 80.00 60.0% urchase Frequency (per 1,000) 70.00 50.0% 60.00 40.0% 50.00 40.00 30.0% 30.00 20.0% Exposur 20 00 20.00 Pu 10.0% 10.00 0.0% - 1 / 1 1 / 2+ 2 / 1 2 / 2 2 / 3+ 3+ / < 3 3+ / 3+ Policy Number Vehicles / Drivers 19 Data Summary – Bodily Injury Limit 20 10
9/28/2011 Model Development 21 Modeling Techniques Neural N Network k Decision Regression Tree Analysis Ensemble 22 11
9/28/2011 Quotes Submitted Analysis 23 Submitted Likelihood Model – Decision Tree 12
9/28/2011 Decision Tree – English Rules � Low Estimate � IF total_pages < 11.5 � AND total_ssl_page < 9.5 AND total ssl page < 9 5 � Quote Initiated: 0.4% � High Estimate IF aig EQUALS 1 � AND 11.5 <= total_ssl_page � AND 4.5 <= visit_time � AND means_of_entry IS ONE OF: WEBMAIL NON-REFERRED SPONSORED SEAR � NATURAL SEARCH NATURAL SEARCH AND total_pages < 15.5 � Quote Initiated: 71.4% � Quote Submitted – Number of Prior Visits Number of Prior Site Visits 1.200 R e 1.000 1 000 l 1.000 a S 0.889 t u i b v m 0.779 0.800 e i t 0.668 L t i i 0.600 0.557 k n e g l i Q 0.400 0 400 h u o o o t d e 0.200 o f 0.000 0 1 2 3 4 Number of Prior Visits Relative Likelihood of Submitting Quote 26 13
9/28/2011 Policy Purchased Analysis 27 Age of Driver Model: Purchase made during session 1.40 18% 131% 16% 1.20 118% 117% 111% 14% 104% 104% 103% 1.00 100% Exposure Percentage 12% Relativity 0.80 10% 76% 72% 69% 67% 8% 62% 0.60 47% 6% 41% 0.40 4% 4% 0.20 20% 2% 0.00 0% Sessions with Completed Quote Age Driver 1 Modeled Relativity One Way Relativity 28 14
9/28/2011 Education Model: Purchase made during session 2.50 50% 45% 2.00 40% 196% 35% Exposure Percentage 155% 1.50 149% 30% Relativity 146% 131% 126% 25% 1.00 100% 20% 15% 70% 0.50 10% 5% 0.00 0% Sessions with Completed Quote Primary Driver Education Modeled Relativity 29 Price Sensitivity – One Session Model: Purchase made during session 1.20 60% 106% 1.00 100% 50% Exposure Percentage 0.80 40% 79% 76% 76% Relativity 71% 0.60 30% 52% 0.40 20% 0 20 0.20 10% 10% 0.00 0% Sessions with Completed Quote Modeled Percent Final Quoted Premium greater than Minimum Quoted Premium Relativity 30 15
Recommend
More recommend
