Four Segmentation Models for Improved Results€¦ · Four Segmentation Models for Improved Results. Kristine Ziccardi. Mike Bryant. 2019 User’s Conference ... promote product A

2019 User’s Conference

Four Segmentation Models for Improved Results

Kristine Ziccardi

Mike Bryant


What you just heard:

Relevant Data Sources

Predictive Data

P$ycle Data

Demographics

Mapping / Visual Data

Risk Data (FICO)

Four Segmentation Models:

1. Value Scoring

2. Lifestage

3. Look-alike

4. Next Product


Data Sources

Segmentation

Channels/Tactics


“What I see is an imperfect understanding of where runs come from!”

Peter Brand – Moneyball


THE TIPPING POINT

Flock of

School of

Swarm of

Herd of



4 Segmentation Models


(LOW RISK/ HIGH POTENTIAL)

(HIGH RISK/ HIGH POTENTIAL)

(LOW RISK/ LOW POTENTIAL)

(HIGH RISK/LOW POTENTIAL)

HIGHO

ppor

tuni

tyRisk

HIGH

LOW

Opportunity definition –Ability/potential to cross-sell into this relationship

Risk definition –Assessing the likelihood of account closure or balance diminishment


SEGMENT 1:

“Analytical approach that leverages information such as profitability, balances, tenure and product mix to help identify members/customers that drive value.”

Value Scoring1

Lifestage2

Look-alike3

Next Product4


0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

1 2 3 4 5 6 7 8 9 10

% o

f Res

pond

ers

Capt

ured

% Marketed

40% Marketed

Score Cumulative % of Households

Responders Captured

1 10% 42%

2 20% 65%

3 30% 83%

4 40% 93%

5 50% 95%

6 60% 97%

7 70% 98%

8 80% 98%

9 90% 99%

10 100% 100%

93% Responders

40% gets you 93%

60% gets you 7%


SEGMENT 1: VALUE SCORING

Value scoring:

Platinum

(6.5% of HHLDs drives

64.5% in profit)

Bronze

(65.3% of HHLDs drives

-54.9% in profit)

6.5% 10.9%17.3%

65.3%64.5% 69.8%

20.5%

-54.9%-60%

-50%

-40%

-30%

-20%

-10%

0%

10%

20%

30%

40%

50%

60%

70%

$0

$10,000

$20,000

$30,000

$40,000

$50,000

$60,000

Platinum Gold Silver Bronze

Value Score Distribution

Avg Deposit $ Avg Loan $ % of HHs Percent of Profit






HIGH

Opp

ortu

nity

Risk

HIGH

LOW

Platinum 2,400 Platinum 4,000

Platinum 800 Platinum 800

Bronze 12,000 Bronze 0

Bronze 20,000 Bronze 8,000

POTENTIAL STRATEGIES?


CASE STUDY: VALUE SCORING

Objective: Understand who the CU’s top households were with the intent of ongoing engagement and appreciation.

Action: Value-scored based on following attributes: Profit, Balance, Unique Products

Strategy: Anniversary & Birthday triggers focused on thanking households for their membership

Results: Average response: 15% Avg ROI: 3,000+%


SEGMENT 2 Lifestage

“Leverages demographic ingredients to provide further visibility into the member/customer based on what financial stage they are in.”

Value Scorings1

Lifestage2

Look-alike3

Next Product4


SEGMENT 2: LIFESTAGE


HIGH





Opp

ortu

nity

Risk

HIGH

LOW

POTENTIAL STRATEGIES


CASE STUDY: LIFESTAGE SEGMENTATION

Objective: Encourage less traditional savers to save at the CU

Action: Identified by lifestage those members most likely to save for a baby, vacation and wedding

Strategy: Email campaign to members within various lifestages having high propensities for having a baby, taking a vacation or paying for a wedding

Results: Response: 2.0% ROI: 1,786%


SEGMENT 3: LOOK-ALIKE

“Learns from those who engage, finds those who fit a similar profile as the performers”

Value Scorings1

Lifestage2

Look-alike3

Next Product4


SEGMENT 3: LOOK-A-LIKE MODEL

Optimizes data to ID HHs with high potentialMaximizes opportunity by

increasing relevance






HIGHO

pp

ortu

nit

yRisk

HIGH

LOW


Sample Strategies

Campaign #1: Leverage color capabilities to promote product A and product B to this audience that has high propensity to sign up for more products.

Campaign #2: This segment is showing propensity to engage in several additional products. Testing various combinations will enlighten us as to what product mix best works for this group.

The Audience• 10% of your customer base• 65% of all your overall revenue• Ave balance: $175,000• # of products: 5.7

Segment Information• 20% have this product• 60% are web members• 75% have valid email on file• 45% have this product

Demographic• Household Income

– 30% - Mid Income $40-$80k– 40% - Low Income Under $40k– 50% - High Income $80k+

• Average Age = 45• 56% are homeowners

(LOW RISK/ HIGH POTENTIAL) (HIGH RISK / HIGH

POTENTIAL)

(LOW RISK / LOW POTENTIAL)

(HIGH RISK / LOW POTENTIAL)

HIGH

LOW

Op

por

tun

ity

Ind

ex

Risk IndexHIGH

PROFILE HIGH RISK/HIGH POTENTIAL


CASE STUDY: LOOK-ALIKE MODEL

Objective: Expand relationships among households with potential

Action: Top HH analysis based on product usage and appended data. Identified less valuable households resembling top HHs

Strategy: Reboarding trigger campaign revamped to target these look-a-like households and implemented to replace traditional reboarding program

Results: 2-month avg response: 4.13% Avg ROI: 1,174%


SEGMENT 4: NEXT PRODUCT

“Art meets science, leverages many of the aspects of the other segmentations, best used in POS channels.”

Value Scorings1

Lifestage2

Look-alike3

Next Product4


SEGMENT 4: NEXT PRODUCT

Product Balance

Lifestage

Geography


Leverages profitability,

tenure, balance and product mix

Leverages demographic ingredients to

provide further visibility into the

member

Learns from those who

engage, finds those who fit a

similar profile as the “performers”

Art meets science, leverages many of

the aspects of the other

segmentations, best used in

POS channels


BILL WALSH – FOOTBALL COACH

Inducted into Hall of Fame 1993

San Francisco 49ers Coach

3 Super Bowl Titles

He is known for inventing what?


ANSWER: WEST COAST OFFENSE


On DeckPROSPECTING

Intelligent Farming Strategies

Jeanine Perrone & Christopher Ruscher

Four Segmentation Models for Improved Results€¦ · Four Segmentation Models for Improved Results. Kristine Ziccardi. Mike Bryant. 2019 User’s Conference ... promote product A

Documents