How to overcome Big Data challenges
How to overcome Big Data challenges
Introducing:
Satya Lakkaraju Citi
Bryon Evans Experian
Natasha Madan Experian
©Experian 3 4/15/2017 Experian Public Vision 2017
1. Environment
2. Demonstration
3. Analytics on Demand
4. Q&A with Citi
Contents
©Experian 4 4/15/2017 Experian Public Vision 2017
Experian’s Big Data platform
©Experian 5 4/15/2017 Experian Public Vision 2017
What are some perceptions of Big Data?
©Experian 6 4/15/2017 Experian Public Vision 2017
Experian’s Big Data platform enables integrated analysis, development, and deployment of focused solutions for our clients
• The initial environment consists of:
– Client analytic environment
• Shared or dedicated
– Batch production
• Prescreen, account review
– TIP updated multiple times per day
• Client “production” environment (PDB)
• Archives
What is Experian’s Big Data platform Enabling a wide set of capabilities
©Experian 7 4/15/2017 Experian Public Vision 2017
An environment that provides clients unprecedented access to Experian’s consumer data
• Fifteen+ years of month-end 100% depersonalized consumer data
– Data covering over 300 million consumers per month
– Industry leading scores, attributes
• VantageScore® 3.0
• Premier Attributes℠
• Trended Attributes
Analytical environment Access to best in class attributes, scores and trended attributes
No costly infrastructure builds
or IT projects needed to use
©Experian 8 4/15/2017 Experian Public Vision 2017
An environment that provides a comprehensive data and tools to enable clients to create scores, attributes and decisions
• Trade, inquiry, public record, trended, and extended data
– Over 1.1 trillion rows of TIP data for big data analytics
• Best-in-class software
– R
– H20
– Python
– Anaconda
Analytical environment Data and tools to develop insights
©Experian 9 4/15/2017 Experian Public Vision 2017
• Easily create, validate and deploy attributes and strategies using Experian tools
• Attribute Toolbox™ (ATB) and PowerCurve® Strategy Management
– Develop and validate attributes in the Experian analytical environment
– Easily deploy in our File OneSM environment
• Ability to deploy in batch coming soon
Analytical environment Development and deployment
©Experian 10 4/15/2017 Experian Public Vision 2017
Analytical environment demo
©Experian 11 4/15/2017 Experian Public Vision 2017
“Information is the oil of the 21st century, and analytics is the combustion engine.”
— Peter Sondergaard Senior Vice President
Gartner Research
©Experian 12 4/15/2017 Experian Public Vision 2017
Analytics on demand
©Experian 13 4/15/2017 Experian Public Vision 2017
Get more, do more
Key topic Description Approach
Score migration
Gauge movement of key score averages and relationship to performance with industry comparisons
• Calculate VantageScore® migration for client portfolio and overall industry
New account vs. existing accounts
How do balances on/off client differ for those that migrate 20+ points?
Benchmarking Develop key portfolio stats
• Coded for generic benchmark to industry
• Evaluate correlation between lines of business (auto, mortgage, card, etc.)
• Perform validations on portfolio slices by geography, industries, product types
Market entry Conduct prospect development review and analyses
• Identify opportunities between lines of business (auto, mortgage, card, etc.)
• Evaluation of new vs. existing accounts, current vs. derogatory, etc.
• Create peer grouping
Wallet share Assess customer relationships with details about their overall spend
• Share of balance and credit limit (basic)
• Consumer spend available through Total Annual Plastic Spend (Experian TAPSSM) and Estimated Interest Rate Calculator (EIRCSM)
• Trade level spend and EIRC allow for more detailed on-them/off-them spend comparison
Incremental share Compare portfolio usage pre- and post-acquisition
• Due diligence for future acquisitions and post-acquisition performance
• Portfolio reference files flag portfolios/trades
©Experian 14 4/15/2017 Experian Public Vision 2017
Reject bias
• Bias created by a creditor’s approach to risk management
• Results in an incomplete picture of the through the door population
• High decline rates with very consistent decision making result in the most problematic bias
Use case 1 Reject inference
Total Applications
Approvals Declines
Not Booked Unknown
Good Unknown
Bad Booked
Declines
Unknown Good
Unknown Bad
©Experian 15 4/15/2017 Experian Public Vision 2017
Use case 1 Reject inference
“If this applicant had been accepted,
how would they have performed?” Bureau inference
Step 5
Step 5
Identify goods
vs. bads and re-
adjust policies
5
Step 4
Track loan
performance of
the new loan 4
Step 4
Step 3
Identify consumers
tagged in step 2; who
open a similar product
loan within 3-6 months
3
Step 3
Step 2
Experian tags
those applicants in
the big data
environment 2
Step 2
Step 1
Client sends
applicant file
of declines.
1
Step 1
©Experian 16 4/15/2017 Experian Public Vision 2017
Use Case 2 Revenue potential
“How much share
did I capture?”
“Identify
opportunities”
“How profitable are
the consumers?”
©Experian 17 4/15/2017 Experian Public Vision 2017
Use Case 2 Wallet share analysis
The Experian Profit Suite
is integral in understanding
the share of wallet for your
consumer base.
Metrics Bank ABC bankcard Share of wallet
VantageScore 3.0 ® 659
On
-bo
ok
Average # of trades 1.2 19.5%
Average credit line $3,813 15.6%
Average revolving balance $1,107 20.7%
Average annual spend $2,745 21.3%
Average effective yield 25.58%
Annual interest $317 30.0%
Estimated interchange $46 21.3%
Off
-bo
ok
Average # of trades 4.96
Average credit line $20,663
Average revolving balance $4,233
Average annual spend $10,167
Average effective yield 4.93%
Annual interest $739
Estimated interchange $170
Experian TAPSSM and EIRCSM data elements
©Experian 18 4/15/2017 Experian Public Vision 2017
Use Case 2 Targeting top of wallet off – Bank ABC spend
Metrics
Consumer top
of wallet trade
Bank ABC (48%) Other (52%)
Ris
k
Average VantageScore® 661 646 670
Delinquency rate (60+ DPD) 1.9% 2.9% 1.3%
Us
e
Average credit line $5,712 $4,044 $6,772
Average revolving balance $1,644 $1,022 $2,040
Avg revolving utilization 29% 25% 30%
Annual spend $7,195 $4,126 $8,877
Reven
ue
Effective yield 18.89% 22.46% 16.89%
Annual interest $456 $333 $527
Estimated interchange $120 $69 $148
Opportunity for increased revenue exists for those existing
customers where the client is not top of wallet …
Which customers should be targeted?
©Experian 19 4/15/2017 Experian Public Vision 2017
Use Case 2 Segmenting the highest spend cards (off Bank ABC)
27,069
$458
$1,396
<$1,000 $1k–$5K $15k+
>$2
0k
<$2
k
$2k–$20k
3316
$33
$1,160
162
$17
$1,084
74,361
$31
$496
6390
$15
$471
92,518
$31
$188
24,323
$14
$148
77,919
$397
$693
57,851
$378
$245
7,003
$58
$1,191
69,974
$60
$566
44,293
$76
$201
6,648
$87
$1,215
35,155
$98
$608
22,385
$133
$203
$5,001–$10k $10,001–$15k
Revenue $ opportunity (50%)
~$25 million
(27,069 customers)
~$59 million
(119,722 customers)
~$26 million
(76,977 customers)
~$670 million
(4.7 million customers)
*Numbers reflect 10% sample size
*Bank ABC consumers with at least 1 open/active trade
#Customers / Avg Interchg / Avg Interest High Revolving balance
High
Sp
en
d
Knowing which accounts have the most revenue potential can be incorporated into existing strategies for maximum effect
Low
©Experian 20 4/15/2017 Experian Public Vision 2017
• Originations vs. existing accounts
• Score migration by: Geography, industry types, product types
• Historical risk performance
• Portfolio stability
Existing accounts
Metrics:
• VantageScore®
• Credit line
• Balance
• Spend and share of wallet
• Interest rate
• Credit usage (transactor, revolver, etc.)
Compare Bank ABC and peer group across several metrics to better understand:
• Strengths
• Area of opportunity
Use Case 3 – Portfolio benchmarking Benchmark Bank ABC against industry peers
659 661 663 663
727 728 729 730
300
400
500
600
700
800
201212 201312 201412 201512
Bank ABC Peer Group
Va
nta
ge
Sc
ore
® 3
.0
0%
10%
20%
30%
40%
2012 2013 2014 2015
Super Prime Prime Near Prime Subprime Deep Subprime
Bank ABC
0%
10%
20%
30%
40%
2012 2013 2014 2015
Super Prime Prime Near Prime Subprime Deep Subprime
Peer group
©Experian 21 4/15/2017 Experian Public Vision 2017
Q&A with Citi
©Experian 22 4/15/2017 Experian Public Vision 2017
Experian contact:
Natasha Madan [email protected]
Byron Evans [email protected]
Questions and answers
©Experian 23 4/15/2017 Experian Public Vision 2017
Share your thoughts about Vision 2017!
Please take the time now to give us your feedback about this session.
You can complete the survey at the kiosk outside.
How would you rate both the Speaker and Content?