SaFi Bank Space : Meiro attributes mapping

Labels

Data Source

Attribute Name

Attributes Description

Compound?

Dimensions
(Compound Attribute)

Is Unique?

Data Type

Time Limit (Define for non-unique attributes)

Remarks

SaFi source

PII

Mixed Sources

Name

Name of the customer

FALSE

TRUE

string

NA

Check if we can split into first and last name or not

schema: customers.customer.snapshot.v1

property: ${customer.profile.firstName} ${customer.profile.middleName} ${customer.profile.lastName}

note: middleName is optional

PII

Mixed Sources

First Name

First name of the customer

FALSE

TRUE

string

NA

Combine across all data sources where customer name is captured (eg. Take the most recent / most frequent value)

schema: customers.customer.snapshot.v1

property: ${customer.profile.firstName}

PII

Mixed Sources

Last Name

Last name of the customer

FALSE

TRUE

string

NA

Combine across all data sources where customer name is captured (eg. Take the most recent / most frequent value)

schema: customers.customer.snapshot.v1

property: ${customer.profile.lastName}

Geographic

Mixed Sources

City

A customer's most frequently tracked city

FALSE

TRUE

string

NA

Combine across all data sources where city is captured. Most frequent value.

schema: customers.customer.snapshot.v1

property: customer.address.city

Geographic

Mixed Sources

Country

A customer's most frequently tracked country

FALSE

TRUE

string

NA

Combine across all data sources where country is captured. Most frequent value.

schema: customers.customer.snapshot.v1

property: customer.address.country

PII

Mixed Sources

Phone Number

Most recently used phone number from all channels

FALSE

TRUE

string

NA

Filter the phone numbers by the list of do not call numbers, data name: DNC Customers

schema: customers.customer.snapshot.v1

property: +${customer.countryCode}${customer.phoneNumber}

note: phone number is not unique (verified/active phone numbers are unique)

PII

Mixed Sources

Email

Most recently used email from all channels

FALSE

TRUE

string

NA

schema: customers.customer.snapshot.v1

property: customer.email (optional)

note: email is not unique

PII

Mixed Sources

Account Number

Most recently used account number from all channels

FALSE

TRUE

integer

NA

??? customerId or mainAccountId (TODO ACCOUNTS)

PII

Mixed Sources

Age

A customer's most recently tracked age

FALSE

TRUE

integer

NA

(customers) we store date of birth instead of age

schema: customers.customer.snapshot.v1

property: customer.profile.dateOfBirth

PII

Mixed Sources

Gender

A customer's most recently tracked gender

FALSE

TRUE

string

NA

schema: customers.customer.snapshot.v1

property: customer.profile.gender

Engagement

Mixed Sources

First Brand Interaction

First interaction on the brand by the customer from all sources

FALSE

TRUE

datetime

NA

TODO ENGAGEMENTS

Engagement

Mixed Sources

Last Brand Interaction

Most recent interaction on the brand by the customer from all sources

FALSE

TRUE

datetime

NA

TODO ENGAGEMENTS

Kafka

Most Recent Subscription

Most recent subscription plan of the customer

TRUE

-Date
-Plan Type
-Plan Value

FALSE

compound

Date should cover both start & expiry date of subscription plan

There are 3 types of subscription plan for our customer to avail of: "All-Star", "Growth", and "Smart". Each plan gives customers X number of transactions per month

TODO SUBSCRIPTIONS

Kafka

Credit Information

Most recent credit information of a customer

TRUE

-Date
-Total Credit Limit
-Credit Balance
-Credit Type
-Credit Score
-Credit Rating

FALSE

compound

Credit score will be done by Advance's risk team, we will get the score from database directly

TODO ACCOUNTS/LOANS

Kafka

Total Balance

Total balance of a customer

FALSE

FALSE

int

TODO ACCOUNTS

Kafka

Average Daily Balance

Average daily balance of a customer

FALSE

TRUE

float

TODO ACCOUNTS

Kafka

Unbanked Status

Is this person previously unbanked or not

FALSE

TRUE

boolean

Yes = unbanked
No = banked

???

Kafka

Date Of Account Creation

Date of account creation by the customer

FALSE

TRUE

datetime

schema: customers.customer.snapshot.v1

property: customer.createdAt or customer.onboardedAt ???

Kafka

Loan Offer

Loan offer & loan term subscribed by the customer

TRUE

-Date
-Loan offer
-Loan term

FALSE

compound

TODO LOANS

Kafka

NPS Score

Most recent NPS score of a customer

FALSE

TRUE

int

NPS stands for Net Promoter Score which is a metric used in customer experience programs.

Currenlty out of the scope.

TODO CUSTOMERS

Kafka

CSAT

Most recent CSAT score of a customer

FALSE

TRUE

int

CSAT is short for customer satisfaction score.

Currenlty out of the scope.

TODO CUSTOMERS

Kafka

Kiosk Location

Location of Kiosk machine visited by the customer

FALSE

FALSE

string

out of the MVP scope

TODO CUSTOMERS

Kafka

Maximum Interest Rate

The max interest rate to offer to a customer

FALSE

TRUE

int

TODO LOANS

Kafka

Maximum Offer

The max credit line to offer to a customer

FALSE

TRUE

int

TODO LOANS

Kafka

Maximum Term

The max term for loan to offer to a customer

FALSE

TRUE

int

TODO LOANS

Kafka

Successful Sale

Whether it is a successfull cross-sell or upsell

FALSE

TRUE

boolean

???

Kafka

All Transactions

All transactions made the customer

TRUE

-Date
-Type
-Amount
-Biller Name

FALSE

compound

TODO TRANSACTIONS

Profile Stitching

Kafka

Advertising ID

All the Advertising IDs obtained from the customer

FALSE

FALSE

string

schema: customers.customer.snapshot.v1

property: customer.id

Profile Stitching

Kafka

Android ID

All the Android IDs obtained from the customer

FALSE

FALSE

string

schema: customers.device_fingerprint.snapshot.v1

where customerId is same as customer.id

property: data.deviceLabel.androidId (optional because of iOS users)

Profile Stitching

Kafka

IDFA

All the IDFA obtained from the customer

FALSE

FALSE

string

schema: customers.device_fingerprint.snapshot.v1

where customerId is same as customer.id

property: data.deviceLabel.idfa (optional because of Android users)

Profile Stitching

Kafka

IDFV

All the IDFV obtained from the customer

FALSE

FALSE

string

schema: customers.device_fingerprint.snapshot.v1

where customerId is same as customer.id

property: data.deviceLabel.idfa (optional because of iOS users)

Profile Stitching

Kafka

IMEI

All the IMEI obtained from the customer

FALSE

FALSE

string

schema: customers.device_fingerprint.snapshot.v1

where customerId is same as customer.id

property: data.deviceLabel.imei

Kafka

Agent ID

Agent ID of a customer tied to a sales agent

FALSE

FALSE

string

Each sales agent should have their own Agent ID.
Agent ID should be found in Genesys. It is possible for 1 customer to have 2 agent IDs attached

out of the MVP scope

TODO CUSTOMERS

Kafka

Prospect ID

All the Prospect IDs assigned to a customer

FALSE

FALSE

string

schema: customers.customer.snapshot.v1

property: customer.id

Kafka

Customer ID

All the Customer IDs assigned to a customer

FALSE

FALSE

string

schema: customers.customer.snapshot.v1

property: customer.id

Kafka

Customer Feedback

All customer feedback given by a customer

TRUE

-Date
-Action
-NPS
-Feedback

FALSE

compound

Action (eg. Ticket resolution, Chat etc...)
NPS (Customer Satisfaction Rating of value 1 to 5)
Feedback (Thumbs Up / Thumbs Down from Customer Satisfaction Thumbs field

NPS stands for Net Promoter Score which is a metric used in customer experience programs.

Currenlty out of the scope.

TODO CUSTOMERS

Kafka

Declined Loan Application

Whether a loan application has been declined for a customer

FALSE

FALSE

boolean

To check kafka database if there are more data that we could use, currently based on the notification that they are sending out, we only have a yes/no answer for this

[Compound Attribute] Loan Application status to explore

• loan type

• date of result

• status (approved/rejected)

TODO LOANS

Kafka

All Consented Uncontactable Phone Numbers

All consented uncontactable phone numbers of a customer

FALSE

FALSE

string

NA

For sales people to have an overview of the list of uncontactable phone numbers

???

Kafka

Completed Onboarding In The Last 1 Month

Has the customer completed onboarding by the agent in the last 1 month

FALSE

TRUE

boolean

To segment groups of customers and reengage them with marketing activities after their 3 months free subscription plan in their onboarding campaign

it could be computed from

schema: customers.customer.snapshot.v1

property: customer.onboardedAt

Kafka

Number Of Top-Ups

Total number of top-ups made by the customer

FALSE

TRUE

int

The attribute is left generic without timeline for now, we can also explore the following attributes

• No. of top ups in the last 1 month

• No. of top ups in the last 3 months

TODO TRANSACTIONS

Kafka

Number of times contacted by a customer service representative

Total number of times contacted by a customer service representative

FALSE

TRUE

int

Data name: Spins

???

SMS Behaviour

Infobip

All SMS campaigns received / opened / clicked

All SMS campaigns that the customer has ever received / opened / clicked

TRUE

-Date
-Campaign name

FALSE

compound

SMS Behaviour

Infobip

Date of Last SMS Campaign Received / Opened / Clicked

Date of Last SMS campaign sent/opened/clicked to the customer

FALSE

TRUE

datetime

One attribute for each action: received, opened, clicked

Infobip

All Viber campaigns received / opened / clicked

All SMS campaigns that the customer has ever received / opened / clicked

TRUE

-Date
-Campaign name

FALSE

compound

Infobip

Date of Last Viber Campaign Received / Opened / Clicked

Date of Last SMS campaign sent/opened/clicked to the customer

FALSE

TRUE

datetime

One attribute for each action: received, opened, clicked

Email Behaviour

Emailing Engine

Subscription Status

Subscription status of a customer

FALSE

TRUE

string

Yes / No / Unsubscribed

Email Behaviour

Emailing Engine

All Email Campaigns Received / Opened / Clicked

All email campaigns that the customer has ever received / opened / clicked

TRUE

Date
Campaign name

FALSE

compound

One compound for each action: received, opened, clicked

Email Behaviour

Emailing Engine

Last Email Campaign Received / Opened / Clicked

Last email campaign name received / opened / clicked by customer

FALSE

TRUE

string

One attribute for each action: received, opened, clicked

Email Behaviour

Emailing Engine

Date of Last Email Campaign Received / Opened / Clicked

Date of last email campaign sent/opened/clicked to the customer

FALSE

TRUE

datetime

One attribute for each action: received, opened, clicked

Email Behaviour

Emailing Engine

Last Email Consent Given

Most recent email consent of a customer

FALSE

TRUE

string

consent, no consent

Email Behaviour

Emailing Engine

Number of Emails Received / Opened / Clicked

Number of emails received / opened / clicked by a customer for all email campaigns

FALSE

TRUE

integer

One attribute for each action: received, opened, clicked

Email Behaviour

Emailing Engine

Email CTR

A customer's clickthrough rate. It is calculated by (Number of unique link clicks / Number of emails sent).

FALSE

TRUE

float

Email Behaviour

Emailing Engine

Email Open Rate

A customer’s open rate. This is calculated by (Number of email opens / Number of emails sent).

FALSE

TRUE

float

Profile Stitching

Appsflyer

Appsflyer ID

List of Appsflyer ID of customer

FALSE

FALSE

string

App Behaviour

Appsflyer

Most frequently viewed product

Most frequently viewed product by a customer

FALSE

TRUE

string

Transactional

Appsflyer

Availed product

Products availed by a customer

TRUE

-Date
-Time
-Product Name
-Product Type

TRUE

compound

6 months

App Behavioral

Appsflyer

Session duration

Amount of time spent in one session by the customer

FALSE

TRUE

datetime

App Behaviour

Appsflyer

Last Login Date

Last login date of the customer

FALSE

TRUE

datetime

App Behaviour

Appsflyer

Number of Logins

Number of times the customer logins into his or her account

TRUE

• Last 7 days

• Last 30 days

• Last 60 days

TRUE

compound

App Behaviour

Appsflyer

App install

App installs of customer

TRUE

• Date

• Campaign

• Adtype

• Adset

• Channel

FALSE

compound

App Behaviour

Appsflyer

App actions

List of actions & pages viewed on the mobile app by the customer

TRUE

• Actions

• Views 1 month

• Views 3 months

• Views 6 months

• Views 1 year

FALSE

compound

App Behaviour

Appsflyer

Customer Onboarding Drop Off

Drop off screen of a customer during customer onboarding

TRUE

• Date

• Time

• Drop Off Screen

TRUE

compound

Drop off type values:Subscription Model
Subscription Plan
Login Screen
Signup Screen
Identity Verification Screen
Facial Verification Screen
More Information Screen

App Behaviour

Appsflyer

Average time spent on mobile app

Average time spent on app of a customer

FALSE

TRUE

float

Web Behavior

Website

Engagement Score (mobile app)

The customer's engagement score based on the events from mobile app

FALSE

TRUE

float

NA

To define which are the events required for the calculation and if there a specific model that they would like to deploy

Web Behavior

Website

Total Pageviews

Total number of pageviews from customer

FALSE

TRUE

integer

NA

Web Behavior

Website

Pageviews per visit

Number of pageviews per visit from the customer

FALSE

TRUE

integer

NA

Web Behavior

Website

Number of Pageviews

Number of pageviews in the last X days from the customer

TRUE

• Page name

• Views in the last 7 days

• Views in the last 14 days

• Views in the lats 30 days

TRUE

compound

NA

Web Behavior

Website

All Pages Viewed

All URLs which customer has ever viewed

FALSE

FALSE

string

3 months

Web Behavior

Website

Last Visit

Date & time of last visit to the website

FALSE

TRUE

datetime

NA

Web Behavior

Website

All Searches made

All keywords which the customer has ever searched

TRUE

-Date
-Keywords
-Results

FALSE

compound

3 months

Technographic

Website

Last Device Type

Last device type used by customer, E.g. Tablet, Desktop, Mobile

FALSE

TRUE

string

NA

Sample values: Desktop / Mobile / Tablet

Technographic

Website

Most Frequent Device Type

Most frequently used device type by customer. E.g. Tablet, Desktop, Mobile

FALSE

TRUE

string

NA

Sample values: Desktop / Mobile / Tablet

Technographic

Website

Last Device Model

Last device model used by the customer

FALSE

TRUE

string

NA

Sample values: iPhone X

Technographic

Website

Most Frequent Device Model

Most frequently device model used by the customer

FALSE

TRUE

string

NA

Sample values: iPhone X

Technographic

Website

Most Frequent OS

Most frequently OS used by the customer

FALSE

TRUE

string

NA

Technographic

Website

Last OS

Last OS used by the customer

FALSE

TRUE

string

NA

Profile Stitching

Website

Google Client ID

All tracked Google Client ID

FALSE

FALSE

string

3 months

Profile Stitching

Website

Facebook Client ID

All tracked Facebook Client ID

FALSE

FALSE

string

3 months

Web Behavior

Website

Entry Page

First page which was visited by the customer

FALSE

TRUE

string

NA

Web Behavior

Website

Exit Page

Last page which was visited by the customer

FALSE

TRUE

string

NA

Web Behavior

Website

First Referrer

The first tracked referrer based on customer's web activities.

FALSE

TRUE

string

NA

Web Behavior

Website

Last Referrer

The last tracked referrer based on customer's web activities.

FALSE

TRUE

string

NA

Web Behavior

Website

Last UTM Campaign / Medium / Source / Content / Term

The last UTM source / medium / campaign / term / content based on the customer's web activities.

FALSE

TRUE

string

NA

one for each: UTM Campaign / medium / source / content / term

Web Behavior

Website

All UTM Campaign / Medium / Source / Content / Term

All UTM source / medium / campaign / term / content based on the customer's web activities.

TRUE

-Date
-UTM Param

FALSE

compound

3 months

one CA for each: UTM Campaign / medium / source / content / term

Web Behavior

Website

Date of first Form Submission

Date of first form that is submitted by the customer

FALSE

TRUE

datetime

Web Behavior

Website

Date of last Form Submission

Date of last form that is submitted by the customer

FALSE

TRUE

datetime

Web Behavior

Website

Number of Forms Submitted

Total number of forms submitted by the customer

FALSE

TRUE

integer

Web Behavior

Website

All Forms Submissions

All forms that were submitted by the customer

TRUE

-Date
-UTM
-Name of Form

FALSE

compound

6 months

Web Behavior

Website

Engagement Score (website)

The customer's engagement score based on the events from website

FALSE

TRUE

float

NA

To define which are the events required for the calculation and if there a specific model that they would like to deploy

QA

Kafka

Count of Event per Source

Total number of events for this source

FALSE

TRUE

Hidden attribute, pre-req for calculating Engagement attributes

QA

Mixed Sources

Number of unique email addresses

Number of unique identifiers for each customer entity

FALSE

TRUE

int

QA

Website

Number of unique Meiro user IDs

Number of unique Meiro user IDs per customer entity

FALSE

TRUE

int

QA

Mixed Sources

Number of unique phone numbers

Number of unique phone numbers per customer entity

FALSE

TRUE

int