Collection Analytic Building Blocks

© 2014 Fair Isaac Corporation. Confidential. This presentation is provided for the recipient only and cannot be reproduced or shared without Fair Isaac Corporation’s express consent.

Collection Analytic Building Blocks

Mary Du PontPrincipal Consultant, Fair Isaac AdvisorsFICO

FICO™ Forum: Debt Manager™ 9 User GroupFairfax, VA | June 4–5, 1014

Agenda

© 2014 Fair Isaac Corporation. Confidential.2

► The Building Blocks of Collection Analytics

► The Data and Tools Needed to Perform the Analysis

► Questions

3 © 2014 Fair Isaac Corporation. Confidential.

Building Blocks of Collection Strategy

Use of RPC to

obtain promises

Most effective use of Calls to get RPC

Most effective use of Dialer and Collector hour to make calls

To increase promises and

reduce roll

To increase RPC

To increase calls

4 © 2014 Fair Isaac Corporation. Confidential.

Dialer Effectiveness

Use of RPC to

obtain promises

Most effective use of Calls to get RPC

Most effective use of Dialer and Collector hour to make calls

Dialer Reports—Per dialer hour effectiveness- make the best use of the dialer

Call Handling Time—Per dialer hour effectiveness- Use the least time per call

RPC Rate by Time and Day—Call at the right time

5 © 2014 Fair Isaac Corporation. Confidential.

Where to Call

0.00

2.00

4.00

6.00

8.00

10.00

12.00

14.00

16.00

RPC/SYSTEM HOUR

Bucket 1 | Mobile | Weekday Bucket 1 | Home | Weekday Bucket 1 | Work | Weekday

Bucket 1 | FreeDial | Weekday Bucket 1 | Mobile | Weekend Bucket 1 | Home | Weekend

6 © 2014 Fair Isaac Corporation. Confidential.

Collector Effectiveness

► Focus on call handling time►Contact AND No Contact►Study by call disposition

►Set disposition level standards►Study by workgroup►Do not mix del levels

0.0

50.0

100.0

150.0

200.0

250.0

300.0

350.0

400.0

RPC Handling Time

man mel

7 © 2014 Fair Isaac Corporation. Confidential.

Dialer/Collector Effectiveness

► Resource as denominator►System hour►Collector hour►Accounts

Dialer Statistics 1-30 days 31-60 days 61-90 days Total

resource= acct calls per acct

agent calls per acctdialer calls per acct

rpc per acctrpc no promise per acct

rpc promise per acctleft message dialer per acct

left message collector per acctleft message per acct

promise to pay per acctno answer per acct

resource= system hour calls per system hour

agent calls per system hour dailer calls per system hours

rpc per system hour promise to pay per system hour

no promise to pay per system hourdialer message left per system hour

collector message left per system hourno answer per system hour

resource= collector houragent calls per collector hour

rpc per collector hourrpc promise tp pay per collector hour

rpc no promise tp pay per collector hourcollector left message per collector hour

collector no message left per collector hour

8 © 2014 Fair Isaac Corporation. Confidential.

Account Level Penetration Study—Know when to stop call and move to

different treatment

Monthly Penetration Rate Study—Cease calls when penetration no longer

increases RPC

RPC Prediction—Call accounts likely to be RPC

When to start calling

Call Effectiveness

Most effective use of Dialer and Collector hour to make calls

Use of RPC to

obtain promises

Most effective use of Calls to get RPC

9 © 2014 Fair Isaac Corporation. Confidential.

► Strategy Questions►When to start calling

►collection entry strategy►When/where to call

►RPC rates►How often per day to call

►Penetration rates►When to stop calling and try

something else►When the odds of reaching the customer

on the next call falls below threshold

Managing Calling Resources

► Strategy Answer►Begin calls at 42 days delinquent►Best time to call is mobile

afternoons►2 attempts per day on weekdays►Stop predictive calling by day 93

when RPC rate drops below 2%

For Every Account!

10 © 2014 Fair Isaac Corporation. Confidential.

When to Start Calling

► When the roll rate is above x

Day 1

Day 2

Day 3

Day 4

Day 5

Day 6

Day 7

Day 8

Day 9

Day 10

Day 11

Day 12

Day 13

Day 14

Day 15

Day 16

Day 17

Day 18

Day 19

Day 20

Day 21

Day 22

Day 23

Day 24

Day 25

Day 26

Day 27

Day 28

Day 29

Day 30

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

Roll Rate by Risk Band

11 © 2014 Fair Isaac Corporation. Confidential.

When and Where to Call

► RPC analysis to predict which accounts (not call patterns) have greater odds of RPC►Does risk band correlate to contact rate?►New characteristics uncovered that can be used in to further improve

segmentation►Days since last customer transaction or Worst cycle in the last 12 months►Count of months delinquent in last 12 months►Customer age—Under 25 population has 400 basis point diff from 25-35

12 © 2014 Fair Isaac Corporation. Confidential.

0%

5%

10%

15%

20%

25%

30%

0

20000

40000

60000

80000

100000

120000

< 28 29-42 43-56 > 56

Acco

unt V

olum

e

Days since last customer activity

DistributionRPC Rate

When and Where to Call

13 © 2014 Fair Isaac Corporation. Confidential.

10.0%

15.0%

20.0%

25.0%

30.0%

35.0%

0

11000

22000

33000

44000

55000

NoScore

201-250

251-300

301-350

351-400

401-450

451-500

501-550

551-600

>600

RPC

Rate

Acco

unt V

olum

e

Collections Score

Distribution

RPC Rate

When and Where to Call

14 © 2014 Fair Isaac Corporation. Confidential.

Call Efficiency

► How often per day to call►RPC rates are highest in early days and early passes►Over 25% of all contact occurs on first pass on first day►Over 40% of all right party contact occurs on first pass on first day

RPC RATE Contact Distribution RPC Distribution

No. Day Pass 1 Pass 2 Pass 3 Pass 4 Pass 5 No. Day Pass 1 Pass 2 Pass 3 Pass 4 Pass 5 No. Day Pass 1 Pass 2 Pass 3 Pass 4 Pass 5Day1 41.3% 25.7% 15.5% 9.1% 6.0% Day1 28.2% 7.0% 2.4% 1.1% 0.6% Day1 47.5% 7.4% 1.5% 0.4% 0.1%Day2 32.4% 20.7% 11.6% 6.2% 5.5% Day2 14.0% 4.0% 1.4% 0.6% 0.3% Day2 18.5% 3.4% 0.7% 0.2% 0.1%Day3 25.0% 15.1% 9.2% 4.5% 3.1% Day3 7.8% 2.7% 1.1% 0.5% 0.2% Day3 7.9% 1.6% 0.4% 0.1% 0.0%Day4 18.9% 11.4% 6.1% 4.0% 2.8% Day4 4.6% 1.9% 0.8% 0.4% 0.2% Day4 3.6% 0.9% 0.2% 0.1% 0.0%Day5 15.0% 9.0% 4.6% 1.9% 3.8% Day5 3.0% 1.4% 0.6% 0.3% 0.2% Day5 1.8% 0.5% 0.1% 0.0% 0.0%Day6 11.0% 7.3% 3.0% 2.2% 0.8% Day6 2.0% 1.0% 0.5% 0.3% 0.2% Day6 0.9% 0.3% 0.1% 0.0% 0.0%Day7 9.7% 5.1% 2.7% 2.6% 0.5% Day7 1.4% 0.8% 0.4% 0.2% 0.1% Day7 0.5% 0.2% 0.0% 0.0% 0.0%Day8 7.1% 4.1% 3.6% 2.7% 2.1% Day8 1.0% 0.5% 0.3% 0.2% 0.1% Day8 0.3% 0.1% 0.0% 0.0% 0.0%Day9 5.0% 4.8% 2.2% 3.0% 2.3% Day9 0.7% 0.4% 0.2% 0.1% 0.1% Day9 0.1% 0.1% 0.0% 0.0% 0.0%Day10 2.8% 2.1% 1.4% 2.3% 0.3% Day10 1.6% 1.1% 0.7% 0.4% 0.2% Day10 0.2% 0.1% 0.0% 0.0% 0.0%

15 © 2014 Fair Isaac Corporation. Confidential.

How Many Time to Call

► If there are 28 days to call, how many do you need to try before it is time to give up?

Sunday Monday Tuesday Wednesday Thursday Friday Saturday

1 2 3 4 5 6 7

8 9 10 11 12 13 14

15 16 17 18 19 20 21

22 23 24 25 26 27 28

✓✓✓✓

✓

16 © 2014 Fair Isaac Corporation. Confidential.

Call Efficiency

► When to stop calling and try something else►Once 50% of week hours

have been called, there is 60% probability of “No Contact”

17 © 2014 Fair Isaac Corporation. Confidential.

Call Efficiency

► Once 40% of week hours have been called, there is >60% probability of “No Contact”

18 © 2014 Fair Isaac Corporation. Confidential.

RPC Effectiveness

RPC/Non RPC Roll Rate Analysis—Which RPC events change rolls and

by how much?

Most effective Use of Dialer and collector hour

Most effective Use of RPC

Most effective

use of RPC

19 © 2014 Fair Isaac Corporation. Confidential.

Measuring Call Effectiveness

► She was going to pay anyway!

► This customer wont pay even if they come to the phone!

► This customer will pay if you can help him understand how/when

20 © 2014 Fair Isaac Corporation. Confidential.

The Value of Contact

► Roll Rate discerns the value of RPC►Typically there are very high

and very low risk accounts for whom RPC has little value (change in behavior)

►Requires granular risk segmentation

Very High High Medium Low Very Low0%2%4%6%8%

10%12%14%16%18%20%

RPC versus No RPC LM versus No RPC LM versus RPC

90 Day Del RPC Roll Rate

LM Only Roll Rate

No Contact

Roll RateVery High 85% 87% 89%

High 80% 84% 91%Medium 70% 80% 88%

Low 50% 56% 62%Very Low 35% 36% 37%

21 © 2014 Fair Isaac Corporation. Confidential.

► Scores!!►No analysis can be interpreted correctly without risk based segmentation

►Scores can be behavior scores, roll rate scores, or payment projection scores

► Data which captures collection and dialer activity►DM9 data►Roll event►Cycle based data is best especially in high volume bucket one

What Tools are Needed?

22 © 2014 Fair Isaac Corporation. Confidential.

Summary

► Ask the right (simple) questions► One question at a time► In the right order► Use the data you have

© 2014 Fair Isaac Corporation. Confidential. This presentation is provided for the recipient only and cannot be reproduced or shared without Fair Isaac Corporation’s express consent.

Thank YouMary Du Pont847-308-8492marydupont@fico.com

FICO™ Forum: Debt Manager™ 9 User GroupFairfax, VA | June 4–5, 1014