Skip to main content
Skip table of contents

Glossary

This topic provides information about the concepts and terminology that you’ll see throughout the Redpoint CDP documentation.

Aggregate

The calculations specific to core and each vertical that define the data that resides in the summary tables and is available for segmentation; the aggregates all have specific data elements passed in feed layouts required for processing. For example, Individual Golden Record (IGR) or Transaction Detail Summary (TDS).

Base tables

The main tables in the data base that contain the base values used for aggregate processing.

Campaign

Activation workflows executed in Redpoint Orchestration that utilize data from the CDP layer for targeted marketing. These campaigns consist of triggers, audiences and segments, resolution, and suppression rules. For example, an offer for 25% off or a birthday email.

Cap value

A cap value allows the user to limit the number of records pulled in a given split either by volume (a set number, like 500) or percentage (for example, only 10% of the total selection for that split). Typically, this is done in the context of A/B testing or where there is limited inventory (for example, I have only 500 of these super-special coupon codes, so I want to limit the selection to 500 people).

Customer Data Platform (CDP)

The Customer Data Platform (CDP) is a software solution that collects and organizes customer data in a single accessible source that pulls from many sources to build the unified records. Specifically, this software is designed to collect, integrate, manage, and store customer data from disparate internal and external systems, sources, and applications (e.g., CRM, ESP, POS, etc.).

Once organized, data is used to create a unified view of the customer that can be used for insights, analytics, reporting, and to activate or orchestrate relevant omnichannel experiences and marketing campaigns. The benefit of deploying a CDP is that it helps improve personalized interactions with individual customers.

Data Definition Language (DDL)

Data Definition Language (DDL) is a standard for commands that define the different structures in a database. DDL statements create, modify, and remove database objects such as tables, indexes, partitions, and users. Common DDL statements are CREATE, ALTER, and DROP.

Data inventory

A comprehensive list of the data required to construct the CDP. For example, PII components and transaction detail data.

Data model

A data model is a visual representation of solution’s database:

  • Tables and the relationships between them

  • Data elements, their data types and length

  • Primary, foreign, and alternative keys

For simplification, data models are divided by logical subject areas, each represented by a separate ERD (entity–relationship diagram). By helping to define and structure data in the context of relevant business processes, data models support the development of effective information systems. They enable business and technical resources to collaboratively decide how data will be stored, accessed, shared, updated, and leveraged.

Core data model

The core data model contains objects grouped in logical subject areas that are not specific to any industry vertical and can deployed on their own, such as Identity Resolution, Customer, Contact Authorization, Campaign & Response Event data, etc.

Industry vertical data model

The industry vertical data model contains objects grouped in logical subject areas that are specific to an industry vertical, for example:

  • Retail: retail product, location, sales, orders, returns, price adjustments, etc.

  • Healthcare: medical/ dental insurance product, provider, healthcare location, appointments, encounters, billing, etc.

  • P&C (Property and Casualty) Insurance: policy, quote, claim, etc.

  • Hospitality: reservation, stay consumption, property, rates, etc.

Database quality metrics

Validity, Timeliness, and Completeness metric colors are determined by set performance values. If the value of a metric is at or above the threshold value, a green oval is displayed. If the value is below the threshold value, a red oval is displayed.

Extension tables

Tables where additional data can be ingested. This data can be exposed in orchestration, but is not part of the aggregate processing without customization.

Feed

A feed is a logical grouping of data in an accepted format (e.g., file, database table, queue, etc.). Feeds are subsets of sources particular to a data grouping. A feed is the atomic unit of Redpoint CDP, the smallest unit of data that the CDP handles.

  • Think of a feed as:

    • Something coming out of a source.

    • A set of data being read, possibly processed, and written someplace else.

  • Examples include:

    • A database table.

    • A flat file generated by a system.

    • A script that reads a table from one location, does some processing on the data, and writes the processed data to another location.

Feed status

Feed status is a process-generated value based on the last instance of a feed run. Possible values are:

Feed Status

Description

Completed

The last feed run executed successfully.

With Errors

The last feed run failed.

Running

The feed is currently running.

Waiting

The feed is in the queue to be executed.

Feed state

A feed state is explicitly set by a user.

Feed State

Description

Active

The feed is ready to receive data.

Paused

The feed is not ready to receive data.

Requested

The user has requested that the feed be created.

Disabled

The user has requested that the feed be disabled.

Feed staleness

A feed is stale if the next feed run has not started within the specified frequency. This is calculated on the fly based upon the Start Datetime of the latest Complete Feed Run for a feed.

There are two causes for a stale feed:

  • A file has not arrived within the determined period of time. For example, on a daily run this would be 30 hours past the Start Datetime of the latest Complete Feed Run. This means a file has not been received in one day + six hours.

  • A file was received but failed to load, causing the feed run to throw an error. It may have thrown an error multiple times (multiple feed runs have ended in error). Since staleness is calculated against the latest Complete Feed Run, this feed would become stale when the determined time period has passed.

Feed frequency

Feed frequency is a value assigned to a feed that sets the time interval between feed run starts. If a feed starts execution after its feed frequency has expired, the feed is stale.

Feed Frequency

Description

Hourly

1 hour + 30 minutes after the Start Datetime of the latest Complete Feed Run. If a feed has a Feed frequency of Hourly and the next feed run has not started within 1 hour + 30 minutes after the Start Datetime of the latest Complete Feed Run, the feed is stale.

Daily

1 day + 6 hours after the Start Datetime of the latest Complete Feed Run. If a feed has a Feed frequency of Daily and the next feed run has not started within 1 day + 6 hours after the Start Datetime of the latest Complete Feed Run, the feed is stale.

Weekly

1 week + 1 day after the Start Datetime of the latest Complete Feed Run. If a feed has a Feed frequency of Weekly and the next feed run has not started within 1 week + 1 day after the Start Datetime of the latest Complete Feed Run, the feed is stale.

Monthly

1 month + 3 days after the Start Datetime of the latest Complete Feed Run. If a feed has a Feed frequency of Monthly and the next feed run has not started within 1 month + 3 days after the Start Datetime of the latest Complete Feed Run, the feed is stale.

Yearly

1 year + 3 days after the Start Datetime of the latest Complete Feed Run. If a feed has a Feed frequency of Yearly and the next feed run has not started within 1 year + 3 days after the Start Datetime of the latest Complete Feed Run, the feed is stale.

Feed run quality metrics

Feed run quality metric

Description

Validity

Number of records processed compared to the number of records quarantined.

Completeness

Number of records processed compared to the average number of records processed in the last 10 feed runs.

Timeliness

The amount of time a feed run is late (expressed as a percentage).
Examples:
100%--Feed run executed on time (for example, an hourly feed completed in less than six minutes after its scheduled run time).
50%--Feed run was completed halfway between its scheduled run time and the next scheduled run time (for example, an hourly feed completed 30 minutes after its scheduled run time).
0%--Feed run executed more than an hour after its scheduled run time.

Feed layout

A standard file layout that is used for data ingestion. Each feed layout is designed to capture data related to the specific business subject area, such as customer profile, loyalty account, retail product, retail transaction, etc. A feed layout contains a list of standard data elements that should be consumed as well as the mapping of each data element to the table and column within the database along with the transformation rules, validation rules, data population rules, and valid values.

  • Core feed layouts: Standard file layout used for core data ingestion that is not specific to any industry vertical.

  • Retail vertical feed layouts: Standard file layout used for data ingestion that is specific to the retail industry vertical.

Refer to Data ingestion basics and Data ingestion details for more information.

Format validity

Address

Redpoint CDP:

  • Standardizes and formats addresses according to local postal standards.

  • Corrects and completes addresses where data are incorrect and/or incomplete.

  • Provides a verification code that indicates one of the following:

    • An exact match to the postal database was found.

    • Multiple matches were found.

    • No match was found.

For US addresses, Redpoint CDP provides a reason why an exact match was not found (for example, a missing suite number).

Phone

Redpoint CDP uses a third-party phone format validator to determine if the number is valid. Note that the validator does not determine if the number is in service.

This format is based on E164.

Email

For Redpoint CDP, the valid email format is user@domain.

Redpoint CDP:

  • Checks that the user is 64 bytes or less and meets the documented standards of a username (allowed characters).

  • Checks that the domain is 255 bytes after the @ character.

The overall length of the email address may not exceed 320 bytes.

Ignore

This term refers to records that have custom rules to "ignore" them during data processing. Such records may be perfectly valid but the customer does not want them loaded. For example, an ignore rule could be attached to a set of records such that if the DOB field indicates that a customer is under 18, the associated record will not be loaded.

Lookup tables

Tables in the CDP that contain all of the lookup values.

Lookup values

The static values that are part of the aggregate calculations.

Matching

Match candidates

The matching process is predicated on initially finding match candidates in the existing CDP universe (possible matches) for the incoming records. Frequently this number will be significantly higher than the number of input records depending on the quality and consistency of the input data.

For example, an incoming record with just a first initial, last name, and email address may have more match candidates than a record with a full first and last name, street address, and email address.

Match field

A match field indicates that the attribute provided will be used in the Identity Resolution process.

Data quality scores

Data quality is the process of assessing, monitoring, and improving the fitness of an organization's data for business use. Fitness is generally a measure of the accuracy, completeness, accessibility, and timeliness of data. Data quality includes capabilities for processing data (integration, parsing, cleansing, matching, and so on) and for assessing data (monitoring, profiling, reporting). Data Quality is also closely related to data governance (metadata, security / privacy / compliance, performance, and so on).

Data quality is used by an organization to understand and improve outcomes that depend on high-quality data, such as customer experience or supply-chain management.

Data quality address score

See address hygiene valid format for an explanation of the term.

  • The data quality address score is calculated when data quality is calculated during feed ingestion.

  • The data quality address score is the address hygiene valid format percent (the percentage of records that have a valid address format).

Data quality phone score

See valid phone format for an explanation of the term.

  • The data quality phone score is calculated when data quality is calculated during feed ingestion.

  • The data quality phone score is the phone hygiene valid format percent (the percentage of records that have a valid phone number format).

Data quality email score

See valid phone format for an explanation of the term.

  • The data quality email score is calculated when data quality is calculated during feed ingestion.

  • The data quality email score is the email hygiene valid format percent (the percentage of records that have a valid email format).

Data quality score

  • The data quality score is calculated during feed ingestion.

  • The data quality score is a measure based on percentage of verified addresses and valid email/phone format.

Parent-child relationship

  • source = parent

  • feed = child

  • table column = grandchild

Personally identifiable information (PII)

Personally identifiable information (PII) is any data that can be used to specifically identify an individual. On the Feed Definition page, individual Feed Columns have a PII section that indicates if the field contains PII (Y or N).

Priority

When you define a source, you set the source's display priority, 1 (highest priority) to 999 (lowest priority). When source names are displayed by priority (such as on the CDP Summary chart) they are listed by priority number (highest priority number to lowest priority number).

Quarantined

  • Quarantined means a record had one or more validation issues and has been removed from the process and written to a file in the quarantine directory.

  • If the incoming feed has duplicates based on the defined primary key, the "extra" records are removed from the process, counted, and reported via the UI.

  • Records that are removed from the process based on a business rule are ignored.

Reference data

Used to ingest lookup values that are specific to the customer (Business Unit, Brand, Payment Type, Sources, etc.). These values will be included in feed layout validation rules, loaded into appropriate lookup tables, and might require business rule adjustments. 

Source

A source is a way to group and categorize feeds (i.e., a source system); it is comprised of one or more feeds. A source could consist of a single feed or a hundred feeds.

For example:

  • Customer operational systems (POS, e-comm, etc.)

  • Third-Party (e.g., CRM, social media, demographic append, etc.)

Source status

This value is an indicator of how well your source is performing. The status of an individual source is based upon the statuses of its included feeds.

Source status

Description

Good

All feeds included in this source executed successfully within their scheduled feed frequency.

Error

One of the feeds included in this source has a feed status of With Errors.

Stale

One of the feeds included in this source is stale.

Source to feed layout

The process during which all required data attributes are mapped from their source to a relevant feed layout. This is the first deliverable in the onboarding process. For example:

  • Mapping PII to Party_Profile feed layout

  • Mapping store data to a location feed layout and relevant data required to its extension table

Refer to Data ingestion basics for more information about this process.

Split

Splits allow you to edit your audience down to your essential customers. A split is a predefined segment added to the audience during audience definition. A split excludes all customers except for the customers defined by the segment. In essence, it "splits" off the group of customers defined by the split segment.

Summary tables

The tables where the data summaries reside.

Suppression

Suppressions allow you to edit your audience down to your essential customers. A suppression is a predefined segment added to the audience during audience definition, but in this case the segment rules exclude customers from the audience.

System

System is a place that holds data at rest or data records at creation.

For example:

  • A CDP

  • A data warehouse

Table metadata

Information about table schema and names, column names, datatypes, nulls, primary, foreign, alternate keys, referential integrity constraints, etc.

Table metrics

The table metrics are calculated from all the feed runs using the following time values and table change counts:

Table metrics time values and table change counts

Description

Start Time

Last completed feed run time (24 hours).

End Time

Last completed feed run time.

Inserted Count

Number of records inserted into the table.

Updated Count

Number of records updated in the table.

Deleted Count

Number of records deleted from the table.

Target

A target is where the data are written.

Valid ranges

Valid ranges reference a range of values acceptable for a given attribute. For example, an Age attribute may contain a range from 0-100, and any value outside that range indicates a problem.

Valid values

Valid values are the values allowed for a particular column (defined in the lookup table). A value outside of those defined in Valid Values would result in an error. Valid values reduce data entry errors and help maintain consistency/data integrity.

JavaScript errors detected

Please note, these errors can depend on your browser setup.

If this problem persists, please contact our support.