The Revenue Mindset Shift Addressing false positives Sam Hartung Whitepages Pro, Partnership Risk Manager
The consumer expectation shift Convenience Curated shopping experience Speed Digital world with fast delivery Cross-border Cross border connectivity 2
Cost centers and sales killers 15% Increase in fraud spending year over year 3
The players Friction Opportunity MERCHANT ACCOMMODATIONS Globalization Mobile Device Usage BOPIS Same-day Delivery Easy Authentication Low-friction Purchasing 4
The balancing act FRAUD PREVENTION CLEARING GOOD ORDERS QUICKLY 5
The revenue mindset shift Easy authentication Mobile device usage Low-friction purchasing Globalization of product lines 6
The risk of not evolving 33% of declined transactions are false positives $118B annually in lost revenue incorrect declines Source: https://www.javelinstrategy.com/press-release/false-positive-card-declines-push-consumers-abandon-issuers-and-merchants 7
The impact to lifetime value 8
A holistic approach to fraud prevention
The way to improve sophistication LAYER 4 Network data LAYER 3 Identity data LAYER 2 User behavior LAYER 1 Device ID 10
Sophistication in action LAYER 4 Network data Kount Boost Rating LAYER 3 Identity data IP is Proxy + Recently Seen Email LAYER 2 User behavior Unknown User LAYER 1 Device ID 11
= good order = bad order 12
= good order = bad order Identity data A look at just IP proxy and recently seen email Plenty of bad orders, but also plenty of good customers Can we help move them out of review? 13
= good order = bad order Network data Adding Boost Rating gives more fine tuning Update the existing review rule to order is >$100 and no email to name match? 14
= good order = bad order Behavior biometrics Let s only review unknown users which we can detect via Behavior Biometrics 15
= good order = bad order Modifying rules 16
More review rule examples Device ID + Identity Data Network Data + Identity Data Device ID + Identity Data Recently cleared cookies + Identity Check Confidence Score Fraud Score < 50 + Email First Seen < 15 IP Proxy + Billing Address does not match IP state + Not Mobile device 17
Measuring false positives Three main approaches to increase effectiveness and cost: Call center tracking Requires human time for outbound efforts Hard to aggregate and quantify written notes Re-analysis by Sr. Agents Expensive to dedicate resources to task Requires pulling most experienced agents off of daily queue Control group accepting a subset of declined transactions Requires a tolerance for increased chargeback rate Needs detailed planning a controls in place 18
Whitepages Pro
Our approach to global data 20 years of data experience We meet every vendor in person Sophisticated data science Powerful linkages Trusted data quality, coverage, speed and accuracy in the marketplace 100+ global data sources, including real-time and proprietary data Rigorous, consistent vetting of our machine learning and algorithms to optimize performance Curated 1M+ linkages a day of unregulated data to verify identities 20
Global data, global scale 5B global contacts 2.5B emails 8B global data links 790M 380M phones addresses 21 *not to scale
Whitepages Identity Graph A living global graph Email Person Phone Address IP 8 billion global identity records 22
Global Identity Check API Determine if name, address, phone, email, and IP data belong to the same customer Name Phone Address Email IP Address Linked to email Is valid Is valid Is valid Is valid Linked to phone Line type Diagnostics Disposable email Is proxy Linked to address* Carrier Linked to phone Auto-generated email Distance from address Confidence Score Country code Linked to address Linked to name Linked to address Resident name* Age range of resident First seen date Domain creation date Registered name Distance from phone Connection type Geolocation Single score 0 to 500 Subscriber name Is prepaid Age range & gender of subscriber Is commercial Receiving mail Delivery point Is commercial Is forwarder Linked to name Age range of registered name Global data element US and Canada only * Availability of this data attribute is limited to certain geographical areas. 23
Identity Check API and our Confidence Score Utilizing Identity Check data and our Whitepages Pro Identity Network Score is powered by Identity Check API data + Whitepages Pro Identity Network Real time identity prediction Usable on both ends of the spectrum Our algorithms Complex data + Identity Network Score Incorporates industry physical goods, digital goods, travel, online lending, marketplaces 24
Q&A Learn how we work with retailers https://pro.whitepages.com/industries/retail/ Explore our blog https://pro.whitepages.com/blog Follow us @WhitepagesPro Questions on today s presentation? Contact Sam Hartung: shartung@whitepages.com 25
Thank you 26
1 Understand your business 27
Fraud trends Customer Profiles Marketing Programs Fraud Strategy Current metrics Goals 28
2 Take a layered fraud approach 29
3 Treat your fraud team like front-line sales 30
What fraud teams worry about Accurately identify good orders, prioritizing customer satisfaction Improve overall customer experience Quickly draw conclusions to handle peak volumes Create resilient verification process to detect fraud Effectively manage number of orders in the queue Master difficult decisions with complex orders 31