CDP Glossary

A composable CDP is an agile configuration of component solutions that enables full usage of your data and MarTech stack to help you achieve your unique use cases. The goal of composability is to put things together in a balanced way that lets you do your job more easily, at the lowest possible cost, with the most accuracy, fewest impediments and the least possible duplication.

The core design principles of composability are modularity, autonomy and orchestration, i.e., doing a well-defined function (modularity), doing it independently (autonomous) and in a way that it can be part of a bigger process (orchestration).

Why

An ability to connect to any end system for the collection and storage of data and to bridge the last mile to the customer allow a composable CDP to support any existing or future CX use case. Users maintain the freedom of choice for which applications to connect to, whether an analytics platform, AI or another emerging technology.

How

A composable CDP should still retain what it means to be a CDP, such as solving for database redundancy, performing data cleansing activities, identity resolution and creating a Golden Record at the point of data ingestion. Having the flexibility to choose best-of-breed components does not negate the need to ensure that all data within the CDP – from ingestion to activation – is ready for business use. A composable CDP is also meant to reduce complexity. When there are too many moving pieces to manage, the burden shifts from your software to your people, so buyers should be skeptical of systems that have a separate tool for everything.

Data-in-place describes running a CDP directly on a data cloud vs. a CDP that works indirectly with a data cloud. A CDP that works directly on a data cloud can be a SaaS or run on a private cloud; the distinction with a SaaS is that the data that runs through the CDP remains in the data cloud instance. The customer – not the CDP vendor – controls the database.

Why

Reasons for a data-in-place deployment include flexibility, scalability, convenience and performance advantages. These include the ability to scale resources dynamically, access advanced analytics capabilities, and integrate with a rich ecosystem of complementary services. Enhanced security is another benefit; when the customer controls the database, the customer data that resides in the CDP never has to leave the customer’s security perimeter. Data-in-place means data is not replicated.

How

By using a data cloud as its primary customer database, a composable CDP gives customers complete CDP functionality available as SaaS with zero data replication. When the composable CDP works directly with everything in the data cloud environment – data, AI, machine learning, analytics, BI and reporting tools, the partner ecosystem and end channels – customers shorten time to value, while benefiting from fit-for-purpose data for emerging use cases like GenAI chatbots and others that require a high-quality data foundation.

Data Activation relates to dropping off or passing off fit-for-purpose data where it needs to be to achieve the desired marketing and/or business use case (e.g., personalized CX). Downstream systems might include a personalization engine, an email service provider, a multi-channel marketing hub, an eCommerce platform – any destination that typically interacts with a customer. In this construct, a CDP is said to “activate” a segment, i.e., push an audience and/or metadata to an interaction channel.

Why

By activating data that is ready for business purpose at the precise moment of interaction and driving multi-channel messaging from a single interface, brands engage in the cadence of the customer.

How 

The Redpoint CDP’s out-of-the-box connectors support connectivity to a preferred activation platform with marketing-ready data and closed-loop response tracking.

Data ingestion is a core CDP functionality that describes the activity of bringing data about a customer from multiple systems into the platform.

Why

For a robust, enterprise-grade CDP, data ingestion means ingesting all forms of customer data – first-party, second-party, third-party, batch, streaming, social. All forms and types of data represent signals about a customer, and thus all forms and types should be collected.

How

One factor to consider about a CDP’s approach to data ingestion is the cadence, particularly when thinking about outcomes or use cases that may require a real-time interaction. Continuous, real-time data ingestion from sources and systems that are themselves in real time ensures that data moving through the platform is in the cadence of the customer.

Data quality is the process of accessing, monitoring and improving the fitness of an organization’s data for business use, i.e., making sure that data is accurate, complete, accessible and timely and that it meets the needs of your specific business use cases.

Why:

Completing data quality processes at the point of data entry ensure they only have to be done once, and ensure that “dirty” data is never used for analytics, machine learning, marketing or even for developing an understanding of a customer, household or business entity.

How:

Different CDPs have different approaches toward data quality. Some confuse the term with simply bringing various signals about a customer together, equating ‘data quality’ with after-the-fact identity resolution. In the Redpoint CDP, data quality processes are a core  capability, ensuring trustworthiness of customer data and giving marketers and business users confidence to execute on the outcomes and use cases that depend on high quality data. By contrast, some CDPs claim that data quality is not a core responsibility of the platform, that it can be done downstream by a third-party system, or that it is an overrated feature, i.e., identity resolution will perform some basic data quality processes good enough for most use cases. It’s important to know how data quality is handled in a CDP, and whether it is thorough enough for use cases that depend on having a highly precise unified profile. Duplicate records and a poor CX (irrelevant, untimely, etc.) are among the implications of not giving data quality the attention it deserves.

Data cleansing refers to the process of detecting and correcting disparities and errors within a data set to ensure data is accurate, consistent, complete and that it conforms to an enterprise standard.

Why:

Data cleansing lets an organization know that data intended for marketing and business use is fit-for-purpose. It guarantees:

• Accuracy – A value in a database field should represent the value of a real-world entity, i.e. names are spelled correctly, an address exists exactly as represented, etc.
• Completeness – Data should be entered as expected without missing or unexpected components, i.e. an address includes a street number, a transaction has two digits after the decimal, etc.
• Conformity – All systems should adhere to the same data standards across the organization, i.e. a date entered as DD/MM/YY rather than YY/DD/MM.
• Consistency – All systems should use the same recorded data, i.e. a customer’s lifetime transactions should be reflected as the same amount for any system using the data.

Data quality issues surface for a variety of reasons. Call center associates incorrectly fill out a form, forgetting to collect a customer’s telephone number and instead entering 111-1111. A Gmail account shows .gov instead of .com. John Smyth is listed as John Smith. Any of these or other issues, unaddressed, may lead to poor customer experience. Using the wrong value of a lifetime transaction due to a misplaced decimal, for instance, might inadvertently put a customer in the wrong loyalty tier.

How

The Redpoint CDP data cleansing capabilities include automated, exception-based processing, parsing and standardization tools that cleanse data as it ingested into the CDP.

The process of converting data from one format to another, data parsing is automated in Redpoint using rules and algorithms to identify relevant data elements from various sources and formats.

Data parsing makes sense of unstructured or unrecognized data by converting it into a structured format that can be analyzed. Examples of data parsing include converting an HTML file into plain text or extracting relevant data from email or social media. A form of data standardization, data parsing is an important tool for analyzing all the disparate signals a customer emits. It will help make sense, for example, of the sentiment expressed on a social media post that might be important for how a brand interacts with a customer on the same or a different channel.

Data standardization converts data elements into a consistent, unified format, such as converting an incorrectly entered DD/MM/YY format into MM/DD/YY, or converting N.Y. to NY.

Why:

Data standardization is the bulwark against data nonconformity, making sure all customer data is organized and formatted in a standardized way.

How

Like data parsing, data standardization is automated in the Redpoint CDP; the platform automatically identifies and transforms data from various sources into a unified format, making it easier to analyze and use for marketing and customer insights.

The process of improving or enhancing an existing data set by adding additional information from external sources or by calculating data “aggregates,” data enrichment is an important element in developing a deep understanding of a customer by creating a more robust, trustworthy unified customer profile.

Why

Data enrichment is an important process for keeping up with the cadence of a customer. By including data aggregates such as time on page, images clicked, etc., a brand has a deeper understanding of customer intent and can provide more relevant interactions.

How

Data enrichment is a continual process in the Redpoint CDP. The unified profile is updated in real time with all signals and information about a customer, providing a fuller understanding of a customer throughout an omnichannel customer journey.

Data lineage, alternately referred to as data curation, is the data lifecycle that includes data origins and where data moves over time.

Why

The ability to track, manage and view data lineage helps simplify tracking errors back to the data source and helps debug the data flow process. A function of any CDP, data lineage helps ensure regulation compliance (GPR, CCPA, HIPAA, etc.). Data lineage also encompasses data quality metrics by following the history of data as it is ingested and made fit for purpose. Tracking and measuring data movement allows marketers to understand, detect and minimize raw data issues and address issues that may affect the quality, quantity, accuracy and veracity of customer data.

How

Data lineage, included as a core capability of the Redpoint CDP, adds contextual value to data by showing how a customer record and/or a customer’s associations change over time, including lifetime value, propensity to churn and other metrics.

Data observability refers to visualization tools that allow marketers and business users of customer data to see the processes and result of bringing data together to create a unified customer profile.

Why

Through graphs, charts, Venn diagrams, etc., data observability lets a user know if data is fit for its intended purpose. Full data transparency allows for the continual monitoring of incoming data, matching, merging and other identity resolution processes, giving marketers trust and confidence in the veracity of a unified customer profile.

How

The Redpoint CDP provides full visibility into the underlying customer data so that marketers know whether the outcomes or interactions with customers will work as intended. If there is a problem with the data or the processes involved with creating the unified customer profile, data observability shows where the problem originates supports and quality improvements.

Describes the monitoring of incoming data to understand its quality, including its completeness, accuracy and viability.

Data profiling identifies characteristics and/or values of the data that may be a problem, such as too many identical records, too many outlying records, blank fields and other errors. Data profiling allows marketers and business users of data to discover and fix problems early, preventing the accumulation of customer data debt.

Data stewardship refers to the people, processes or technology with broad responsibilities for data governance, including monitoring the usability, availability and security of data to meet legal and internal governance requirements.

Why

Data stewardship functionality oversees sourcing, cleansing, mastering, publishing and auditing of data entities to include customer/party, product or site. The purpose is to validate that the data representing a corresponding entity are fit for purpose and available to the people and applications that need them. A data steward, or data stewardship technology, is responsible for overseeing the quality, accuracy and completeness of records, in addition to the processes for correcting errors and making changes.

How

An example of the data steward function is the decision about how or when to delete a record, or what to do about a match discrepancy. Effective data stewardship plays a key role in a personalized CX by recognizing and protecting customer data as an asset, and guarding how it is collected, stored and used for the benefit of both the organization and, ultimately, the customer to whom it belongs. Redpoint data stewardship functionality includes tracking the collection, use and storage of all customer data in full compliance with all data privacy regulations.

Data transparency, an offshoot of data observability, refers to the ability to make in-line changes in a no-code environment, made possible through data observability, i.e., the monitoring of incoming data.

Why

Data transparency is important for letting marketers know why, for example, a match was or was not made when completing identity resolution. With full transparency, marketers can act as data stewards, making on-the-fly adjustments without having to write code.

How

In the Redpoint CDP, marketers have complete visibility into identity stitching, sources, feed status, and completeness and latency. Unlike other CDPs that may provide opaque “black box” functions in areas such as identity resolution, Redpoint provides full visibility into the end-to-end process, and marketers can make on-the-fly adjustments (see: tunable identity resolution) without having to write code.

Identity resolution is the process of finding, cleansing, matching, merging and relating all the disparate signals about a customer across the enterprise (MarTech touchpoints, enterprise systems, databases/lakes, etc.) to give you an accurate, actionable, up-to-date view of an individual customer, a household or entity and the relationships therein, such as the makeup of a household or organization and the changes over time.

Why

Identity resolution is used by marketing and other business functions to analyze, deduplicate and relate customer records to provide a consistent customer experience. Most CDP vendors claim to perform some form of identity resolution. A deterministic approach is the most basic, which is the matching of two or more “Unique Identifiers” – components of data that can be attributed to a single user, like email address, phone, or device ID – or matching a previously known combination of signals, i.e., a phone number with an email address. Probabilistic matching uses algorithms to determine whether two records should be matched, such as the probability that 123 Main St. and 123 Main Street are the same record. Many CDPs offer identity resolution as a black box solution, with deterministic and/or probabilistic methods and algorithms that cannot be changed. Others may offer a combination of their own identity resolution with a callout to third-party services (often specialized batch processes) to perform identity resolution.

How

The Redpoint CDP offers tunable identity resolution as a standard platform feature. Tunable identity resolution gives marketers and business users the power to adjust match quality standards depending on the needed level of exactness for various use cases. A use case that involves sending mortgage information or a consumer’s protected health information (PHI) will require a more exact matching process than sending an email about the opening of a new facility. In addition to tunable deterministic and probabilistic matching, Redpoint also performs identity resolution for simultaneous resolution levels, such as household and individual, or an individual and an organization.

A key difference between tunable identity resolution with Redpoint and other CDPs is that the Redpoint CDP relates signals to one another vs. merely stitching together an identity. This is the key to providing a contextual understanding of an individual through an understanding of various relationships.

A complicated shopping cart fulfillment scenario, for example, might include conflicting signals from three or more different customers – a credit card used to make the purchase is linked to one customer, the device ID where transaction occurred is linked to another, and the shipping address is linked to yet another. Redpoint’s advanced identity resolution features let a marketer know how those signals all relate to one another, and tunability allows the marketer to create a relevant interaction in response to the transaction.

Orchestration describes a dynamic, responsive and up-to-the moment process that culminates in a next-best action. In the last mile to the customer, orchestration involves continually updating the real-time unified customer profile, and making the profile accessible to all end channels in the moment of interaction.

Why

Orchestration is important for consistently delivering personalized engagements in any channel and in the context of a customer journey. Orchestration that is based on deep segmentation delivers customer-centric experiences that fuel data-driven journeys and campaigns.

How

Redpoint orchestrates omnichannel personalization in the cadence of the customer via a closed feedback response loop. Dynamic audience segmentation factors in new customer data and response information (updated preferences, actions/inactions, engagements, geolocation, weather, etc.), which then produces content matched to the right customer in the right channel.

Persistent database key management is the process of linking various signals about a customer to a unique, persistent customer ID that contains all interactions persistent to the customer. A customer, in this context, is understood to mean any party to an interaction a business is trying to understand (an individual, business, household, etc.).

Why

Persistent keys are vital for marketers and business users of customer data to understand how a customer is proceeding through an anonymous-to-known customer journey. Persistent key management provides a longitudinal view of a customer’s history with a brand, important for providing the most accurate, consistent and relevant experiences.

Many CDP vendors take the approach of assigning a key based on common identifiers, such as a name and email address or physical address. In the event a customer uses a new/different email, or an email (or physical) address is not exclusive to the customer, this approach creates uncertainty over the identity of the customer.

How

The Redpoint CDP creates rules for attaching a unique customer ID to every incoming signal about a customer, and rules for keeping and/or replacing those attachments over time – as the customer journey evolves. Persistent key management in Redpoint ensures business users of customer data do not randomly lose or gain information about a customer over time because of the vagaries of how a customer identifies themselves or of how the identity resolution process unfolds. It is the process for ensuring that all normal identity resolution operations provide a consistent, trustworthy identity over time in the face of expected change.

Segmentation is the act of dividing customers into cohesive yet distinct and granular groups that can be targeted with relevant messaging, content and offers delivered in the context of an individual customer journey.

Why

Segmentation is a basic marketing strategy for providing customers with relevant experiences. In a marketing use case, if a cohort (women, 25-34) has expressed a strong affinity for a certain product, basic automated segmentation is the process of carving that audience out of your customer database to market to the audience differently than you would another segment. A limitation of basic segmentation is that the segmentation method – age, gender, income, geo, etc. – may not capture what matters or is meaningful to a customer, nor does a static segment (created at a point in time) capture how a customer proceeds through a customer journey, either across channels or in terms of what is relevant to the customer at the moment of interaction. The limitations of basic segmentation can be traced to the fact that a list-based segment begins to decay the moment it is created. Marketers extract and divide an audience, turn the segment into a list and push the list into a channel, where they execute a campaign. But this process does not account for the fact that customer intent, preferences and even life stages are always in flux.

How

The Redpoint CDP performs dynamic segmentation, using a rules-based approach that accounts for the fact that customer journeys are dynamic across multiple online and offline channels. The Redpoint CDP continually updates data from various sources and adjusts audience segments in real-time, moving with a customer or group of customers across omnichannel customer journeys. Built once, segments change on the fly as real-time data is ingested, optimized according to a pre-determined rule that aligns with the business objective.

Dynamic audience segmentation using rules instead of lists ensures that segments are accurate and up-to-date at the moment of engagement, independent of channel. Some of the conditions and variables that move customers into and out of a universal segment include time-based or geo-based triggers, or any change to a unified customer profile.

Dynamic segmentation is customer-centric, focused only dividing an audience according to the attributes that are important and meaningful to understanding the customer at the moment of interaction, as determined by the underlying data. Segmentation in the Redpoint CDP is also intelligent, i.e., it provides marketers with the ability to create complex segments without requiring code to segment an audience any way a marketer would want – multiple resolution levels, aggregates on demand, nesting, data flows, hooks into third-party systems during audience creation, unmatched metadata management, etc. Complex selections can be done without a single line of SQL.

A triggered action is an automated response based on one or a series of calculations, an “if/then” response mechanism for engaging with a customer. A trigger can be in response to an action taken by a customer, or in response to an external event that affects a customer, i.e., an inventory change, a flight delay, severe weather, etc. Common triggered actions include abandoned shopping cart reminders, prescription refill notifications, an alert to a flight or hotel room booking change.

Why

Originating as application logic executed on specific events in a relational database, if/then calculations trigger when a record is added, modified or deleted. As a response to a specific condition, triggered actions are welcome by customers as relevant, timely and informative. Triggered actions are a way for a brand to express empathy with a customer across an ongoing customer journey.

How

Ranging from simple to complex, triggered actions are often based on a high volume of online and offline data, and there are often competing triggers that must be arbitrated. In the Redpoint CDP, triggers are initiated through channels based on the changing state of an attribute.  Set up on an interactive canvas using a rules designer, users can input commands to fire off a trigger or series of triggers.

A unified customer profile – alternately known as a single customer view, Golden Customer Record, or Customer 360 –  is a precise, real-time representation of a customer, household or business entity that, once activated, is the foundation for providing a personalized customer experience.

Why

A unified customer profile is the result of applying continuous data quality processes and advanced identity resolution as data is ingested. Data is combined from any source (website, mobile app, eCommerce platform, POS, social media, CRM, etc.) to form a holistic unified record of a customer and the customer’s engagement with a brand across every touchpoint.

A unified customer view includes behavioral, transactional, demographic and preference data. A customer’s attributes, data aggregations, a full identity graph and a full contact history are processed and updated in real time, providing marketers and business users with all there is to know about a customer and the customer’s relationship with the brand over time.

How

What differentiates the Redpoint unified customer profile from other CDPs is that advanced identity resolution with tunable matching and merging and persistent key management are included as basic features, distinguishing the Redpoint version from a basic identity graph. The unified customer profile created by the Redpoint CDP also includes real-time throughput from data ingestion through to activation, ensuring that the unified view is always up-to-date and consistently in the cadence of a customer journey.

Real-time interactions is the act of interacting at the cadence of the customer across every touchpoint, driving personalized relevance with a contextual understanding of the customer. Powered by a real-time decisioning engine, a hyper-relevant real-time interaction requires marketers to have a real-time view into customer intent; where is the customer on a buying journey, what was the last action or group of actions taken by the customer?

Why

Real-time interaction is a key requirement for meeting consumer expectations for relevance across an omnichannel customer journey. Real-time interaction empowers a brand to maintain a consistent voice across every customer engagement, arbitrating messages, models, content and data across the customer journey.

How

Redpoint provides interaction design and interaction execution tools that power real-time interactions. Dynamic rules-based assets (images, content and messages) are automatically selected at the moment of interaction on every channel. Redpoint offers A/B/n testing capabilities to know which content engages the most users and best meets KPIs.