10/6/2014
Agenda
THE CUSTOMER JOURNEY ARTS Data Model Foundation for Customer Centric Retail Applications and Services
Customer Centricity
Customer as Core Component of Retail Enterprise Value
Merchandise Vs Customer Centered Retailing
Introduction Customer centered retailing Defining customer Understanding the customer journey Retail context for a customer journey ARTS Operational Data Model V7 Support for customer journey ARTS future direction for supporting customer centric retail applications and services
Dimension
MerchandiseMerchandise-centered
CustomerCustomer-centered
Strategy
Sell “best” stuff at the right price
Create best customer experience
People/Culture
Sell to buy: Buyer central actor – get the best deal
Buy to satisfy customer needs, wants & preferences customer as central actor
Key Metrics
Product GMROI period on period
Customer equity growth over customer lifetime
Organization
Buyer- product category silos
Organize around customer segments
Process
Transaction oriented – short tem
Relationship oriented over long term
Merchandising
Push orientation – retailer drives sales
Pull orientation – customer drives sales
Core Consumer-Customer Lifecycle Concepts
1
10/6/2014
Customer Definition
• A Customer is: – An individual or organization (i.e. a Party) – that assumes a role (PartyRoleAssignment) of a Consumer with respect to the retail enterprise – Who purchases a product or service (exhibited behavior – ConsumerConversionState)
Consumer as a Super-type of Customer
• Consumer
A Customer represents one of several consumer states that make up a consumer life cycle
may be
may be a
Party
is inv olv ed in
A PartyRoleAssignment (role) type that represents the association between the retailer and an individual or organization (Party) where the party is a potential, current or ex-purchaser of goods and services from the retailer.
PartyRoleAssignment
Party Ty peCode
Person distinguishes role of
Organization
Consumer is in a s tate defined by
ConsumerConversionState
Customer
defines condition for
ARTS Sample Customer Lifecycle
Customer Shopper Visitor Inactive Customer Prospect Undifferentiated population
ARTS Sample Consumer Lifecycle State Definitions
Prospect
Visitor
Shopper
ExEx-customer
A consumer that is a potential customer and may be reached through advertising, referrals, or identified through acquired data (e.g. mailing list, prospect list, etc.) A Consumer or prospect that walks into a store or lands on a retailer’s web site. A Visitor that stops and examines merchandise in a way that demonstrates a level of interest and potential purchase
Customer
Inactive Customer
A Shopper that completes a purchase A customer that has been dormant for a retailer designated period of time
Ex-customer
A customer who is inactive and, based on retailer defined criteria, will never become active
A Consumer may exist in one and only one state at any instant in time
Generic Retail Consumer-Customer Portfolio - Life Cycle Context Model
Consumer Lifecycle Metrics
Phase 3 Reviews, opinions, rumors, etc.
Influencers
Reviews, opinions, rumors, etc.
Sentiment about retailer Population Phase 1 Aware of retailer
Prospect
Walk in or land on page
Visitor
Stop/Hold Impression
Shopper
Select & Settle
Customer
Attrition
Advertising, promotions, special events customer correspondence, ongoing customer services and other retailer directed conversations with consumers
Customer Outcome: Lifetime Value
Reactivate & Recover
Consumercustomer Lifecycle Model
Acquisition cost Retention and cultivation cost Net revenue
Retailer Conversion Initiatives
Customer Acquisition & Retention Funnel
Prospects
Conversion
Attrition
Population
Inactive Customers
Historical sales Forecast sales over anticipated tenure or retailer designated period
Discounting model
Visitor Shopper Customer Inactive
The red arrows represent CONVERSION EVENTS and mark the state transition of individuals and organizations as they progress from being part of an undifferenitated popoulation to being CUSTOMERS.
Ex Customers
The funnel graphically illustrates the notion of CONVERSION YIELD.
2
10/6/2014
Customer Lifetime Value – Basic Model
Individual Customer Lifetime Value Aggregated into Valuation Tiers
Customer Portfolio Tiers Based on Customer Valuation
Organizing Retail Strategy Around Customer Portfolio
Retailer’s Customer Equity is the aggregation of its customers’ lifetime values Retailer Customer Equity managed as a portfolio Customer portfolio organized into valuation tiers for investment decision making
Hypothetical Unscaled Grading of CLV Segments for Demand Generation Investment
Retention Probability
Allocation Of Marketing and Promotional Resources
Lead
Iron
Copper
Silver
Gold
Platinum
NONE
D
C
B
A
AA
NONE
D
C
B
A
AA
NONE
D
C
B
A
A
NONE
D
C
B
A
A
NONE
NONE
D
D
B
B
NONE
NONE
D
D
C
C
Platinum Gold Silver Copper Iron Lead
ARTS Data Model Support for Customer Lifecycle Modeling & Analysis
Profitability Crude Sample Allocation of Marketing & Promotional Resources AA 40% A 30% B 15% C 10% D 5%
3
10/6/2014
Strategic: Corporate Net Worth
Where ARTS Plays
ARTS Data Model Work Product Support for Customer Portfolio Management
ARTS ARTS ARTS
ARTS
Retail Enterprise Net Worth (Equity)
Customer Portfolio
Customer Equity Customer Lifetime Value
Customer Acquisition Investment
Future Customer Contribution Margin
Customer Historical Contribution Margin
A
Consumer-Customer Lifecycle Measurement & Characterization ARTS Consumer/Customer KPIs Operational: Factors Affecting Customer Contribution to Corporate Net Worth
ARTS Data Model Work Products
Goal Question Metric ARTS Data Warehouse Model ARTS Operational Data Model
ARTS
Retailer Defined Market Retailer Merchandise/Service Categories and Brands
Consumer-Customer Lifecycle Based Information Model Independent Variables that influence customer behavior
Dependent Variables that reflect the results of customer behavior
Retailer initatives to increase net profitability
Behavioral Segments Merchandise Category
Demographic Segments
Brands
Customer Demographic Characteristics
Channels (where, when media for shopping)
Geographic Segments Customer Geographic Characteristics
Customer Behavior
Customer purchase Promotion/Price condition
Psychographic Segments
Consumer-customer Lifecycle Measurement & Characterization
Operational Data Model Data Warehouse Model Goal Question Metric basis for defining KPI Customer KPI’s used as basis for Consumercustomer Lifecycle Measurement & Characterization
Shopping frequency & recency
A
Entities, attributes and relationships to persist
Transaction volume, sizing and value
Relative Customer Value to the Retailer
ARTS ODM V7 Support for Customer Lifecycle
Occasion
Customer Psychographic Characteristics Modeling Method & Probability Distrution Assumptions
ARTS Data Model Work Products & Consumercustomer Lifecycle Support
Consumer identity and characteristics that describe a consumer independent of their observed behavior and retailer actions
Retailer Consumer Lifecycle Model
RelationshipStage inc ludes
Operational Data Model
LocationLevel
defines status of
ConversionState defines target of contains
is loc ated at
Site
is div ided into
Consumer
contains / is contained in
BusinessUnitSite
may be
Party
ConversionGoal defines
Location BusinessUnit
may be a
contains
is desired outcome from
Named, classified consumer behaviors – which are dependent on retailer actions
describes PartyRole
acts in
defines succ ess criteria for
Retailer Supply Chain Design & Execution
Retailer defined desired customer experience
Chain Store Age Survey
Data for ConsumerCustomer Lifecycle Analytics and Reporting
Retailer Competition & its competitive position
Retailer Advertising, Selling, Fulfillment and Delivery Channels
defines pre-condition of
has parent
TypeCode
ConversionBehaviorType
FunctionCode
defines post condition of
PartyRoleAssignment is a
ConversionInitiative
SellingLocation
Conversion Event State Transition
Z
is in a state defined by
Consumer
Consumer states, state changes and specific events (aka conversion events) that triggered state changes
RetailStore behav ior observ ed through
ConversionEvent
influences occ urence of
triggers change in
ConsumerConversionState
Process completes
state changed by
is place of
Unambiguous association between consumer state changes (aka conversions) and retail transactions
is mediated through / mediates
ProcessChannel
mediates ex ec ution of m ediates
iis referred by
CustomerReferral
is one of defines c ondition for
Customer
Visitor Z
Z
TouchPoint (PointOfInteraction)
is used by
Channel
m ay be a /
Z
Z
defines how , w hen and w here
RetailTransaction
refers
is a
KeyCustomer
marks occurence of
is credited w ith is a party to
4
10/6/2014
PartyAffiliationType Cont act Information
Customer Independent Variables
ContactMethodType
PartyType ContactPurposeType
PartyAffiliation PartyRole
SocialNetworkType
PartyContactMethod
Party
WebSite
SocialNetworkService
PartyRoleAssignment SocialNetworkHandle EmailAddress
Z
Consumer Telephone ConsumerConversionState
ARTS Data Warehouse Model V3
Address
Z
Customer Demographic Controlled Vocabulary
GeographicSegment
ReligionType
Z
Party T y peCode
EducationLevel
KeyCustomer
LifeStageType Geogr aphi c Contr oll ed Voc abul ary
RaceType
KeyCustomerGeographicSegment MaritalStatus CustomerPlaceUsageType
AnnualIncomeRange EmploymentStatusType EthnicType
KeyIndividualCustomerCompositeSegment
OccupationType CompositeDemographicSegment Psychographic Controlled Vocabulary
LifestyleTypeCode PersonalityType ValueAttitudeLifestyleType PersonalValueType
CompositePsychographicSegment
Language Health & Diet Controlled Vocabulary
DietaryHabitType
DisabilityImpairmentType Person
CompositeHealthSegment Act ivit y/ Int ere st Controlled Vocabulary
PersonActivityInterest
ActivityInterest
Analytic Directions
Decomposing “Analytics”
Chain Store Age Survey
demonstrates
Independent Customer Attributes Customer Innate Characteristics Demographic Psychographic Geographic Interests & Activities
Demand Stewardship infers
Customer Needs, Wants, Preferences
defines parameters for
New Customer Acquisition
defines primary drivers of
Customer Retention & Cultivation
Customer Recovery
Customer Attrition
Behavior Dependent Customer Attributes
Retailer Strategy Planning & Execution Product & Services
Pricing
Promotion
Place
Customer KPIs and Performance Measures
Customer Behavioral Metrics
Timeliness Credit risk Purchase behavior patterns
Customer Relationship
Affinity analysis
Acts out retailers role in
Retailer-Customer Interaction Observed Retail Transactions
Customer Orders
Unobserved Product/Service Reviews, Surveys,etc
Models
Customer Behavior
Products (by customer segment and as a way to defined customer segments) Pricing (demand elasticity/sensitivity) Market Basket Analysis Cross sell/upsell Cannibalization
Propensity analysis
Responsiveness to conversion initiatives
Net Sales
Channel preferences Shopping time and venue preferences
REVENUE $
5
10/6/2014
Future Direction for ARTS Data Model Customer Subject Areas
Address consumer-customer privacy issues Extend ODM to capture non-transactional customer-retailer interactions Add subject areas support to define, describe and quantify promotion initiatives Extend data model support for capturing social media conversations about retailer Develop more complete sample customer analytics to support forecasting and modeling
Contact Information Tom Sterling
[email protected]
--------------------------------------------------------------------------------- Create View VW_DW3_RFM_BEHAVIORAL_SEGMENT ---------------------------------------------------------------------------------- This sample view presents a way to segment customers based on the recency --- frequency and monetary value of their behavior. The data source for this --- query is the stored summary table DW3_STRD_SMRY_CT_RP_TRN. The view uses --- common table expressions to create three subqueries to handle summarizing --- recency, frequency and monetary value (which we are populating with --- average net margin). Each subquery is documented. The main query uses --- the NTILE function to assign the returned customer summary values to a --- quintile. The quintile values (1-5) represent bins along three dimensions --- which provide the values used to assign customers to RFM behavioral --- segments. ----------------------------------------------------------------------------------drop view VW_DW3_RFM_BEHAVIORAL_SEGMENT Create VIEW VW_DW3_RFM_BEHAVIORAL_SEGMENT as with CT_RECENCY as ( -------------------------------------------------------------------------- Recency is the number of days from the cutoff date since the last --- transaction was completed for the customer. -------------------------------------------------------------------------select ID_CT ,DATEDIFF(dd,MAX(DC_DY_BSN),'2013-07-01') as RECENCY from DW3_STRD_SMRY_CT_RP_TRN where DC_DY_BSN < '2013-07-01' group by ID_CT )
Sample KPI’s like RFM Provide Concepts and Sample Implementation
Technical Detail Slides
,CT_FREQ as ( -------------------------------------------------------------------------- Frequency is expressed as a transaction occurred every FREQ days --- We calculate it for each customer so it can be returned to the --- outer query for assignment to a quintile bucket for RFM behavioral --- classification. -------------------------------------------------------------------------select ID_CT ,MIN(DC_DY_BSN) as FIRST_PURCH_DATE ,COUNT(ID_TRN) as TRANS_COUNT ,FLOOR(DATEDIFF(dd,MIN(DC_DY_BSN), '2013-07-01') / COUNT(ID_TRN)) as FREQ from DW3_STRD_SMRY_CT_RP_TRN where DC_DY_BSN < '2013-07-01' group by ID_CT ) ,CT_MONETARY as ( --------------------------------------------------------------------------- The monetary value used in this sample is the NET MARGIN. This --- could be replaced by NET SALES. Our example, however is congruent --- with the earlier sample for customer liftime value. Note that we are--- using an AVERAGE TRANSACTION NET VALUE because simply summing the --- net values is too closely correlated with frequency and we want to --- draw a distinction. Average is a good indicator of customer spending--- magnitude over time and will distinguish between frequent convenience--- shoppers versus less frequent stock up shoppers --------------------------------------------------------------------------select ID_CT ,AVG(TRN_NET_SLS) as AVG_SPEND from DW3_STRD_SMRY_CT_RP_TRN where DC_DY_BSN < '2013-07-01' group by ID_CT ) select CT_RECENCY.ID_CT
Quintile Value
Sample RFM Classification
Monetary
FREQUENT
HI GH SPENDER
RECENT TIMELY LOSING STEAM LAGGARD
STEADY NEEDS REMI NDER LOSI NG INTEREST SLOTH
ABOVE AVERAGE SPENDER AVERAGE SPENDER BELOW AVERAGE SPENDER THRIFT
Recency (days)
Freq (days)
ID_CT
RECENCY
FREQ
13.00 36.00 122.00 4.00 68.00 51.00 29.00 43.00 91.00 5.00 7.00 97.00 54.00 13.00 15.00 12.00 40.00 7.00 49.00
Frequency
CURRENT
2 3 4 5 Customer
10048 10037 10021 10028 10082 10065 10005 10084 10041 10085 10056 10064 10019 10099 10092 10025 10010 10001 10015
Recency
1
42 34 49 55 46 44 47 44 47 53 57 46 61 42 44 40 50 46 56
$ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $
Monetary ($) Recency Bin (1-5) Freq Bin (1-5) Monetary Bin (1-5) RECENCY_QUI AVG_SPEND FREQ_QUINTILE SPEND_QUINTILE NTILE 309.27 2 1 1 294.00 3 1 1 293.31 5 3 1 279.44 1 4 1 270.92 4 2 1 257.37 4 2 1 255.19 3 3 1 249.66 3 2 1 243.68 5 3 1 243.00 1 4 1 241.47 1 5 1 239.08 5 2 1 235.90 4 5 1 235.84 2 1 1 235.70 2 2 1 235.16 2 1 1 234.89 3 3 1 234.00 1 2 1 233.81 4 4 1
Privacy Challegne
Privacy metadata
Consumer-customer contracts
Assign privacy rules to consumer, customer, worker and other entities Rules at column/selection set level Data usage permissions Date bounded with renewal
Consumer-customer data usage audit and tracking Consumer-customer right to be forgotten
6
