Consumer Data & Marketing Glossary
Plain-English definitions for the identity resolution, audience targeting, and data enrichment terms you will run into while evaluating (or building) a marketing data stack.
45 terms across 7 categories
Identity & Matching
Identity Resolution
Identity resolution is the process of connecting fragmented signals, such as an email address, a cookie, a phone number, or an IP address, to a single verified person. It closes the gap between anonymous activity (a website visit, a partial customer record) and a real, addressable individual in a data graph. Most platforms combine deterministic matching (exact identifiers) with probabilistic matching (behavioral signals) to resolve identity at scale.
Deterministic Matching
Deterministic matching links records using exact, known identifiers, such as an email address, phone number, or login cookie. If the same email appears in two datasets, deterministic matching treats them as the same person with high confidence. It is the most accurate matching method available, but it only works when a shared identifier actually exists between the two records being compared.
Probabilistic Matching
Probabilistic matching infers identity from a combination of behavioral and contextual signals, like IP address, device type, location, and browsing pattern, when no single identifier is available. No individual signal proves identity on its own, but a strong combination of signals raises statistical confidence. Most enterprise identity systems use probabilistic matching to fill in the gaps that deterministic matching cannot cover.
Match Rate
Match rate is the percentage of input records (emails, names, addresses, or website visitors) that a data platform successfully resolves to a verified identity. A higher match rate means fewer records go unidentified. Match rate varies by data type, geography, and how recently the underlying data graph was refreshed, so it should always be evaluated on accuracy rather than the raw percentage alone.
Consumer Data Graph
A consumer data graph is a database that links identifiers, behaviors, and attributes for individual people across sources into a single, queryable structure. Instead of storing one flat record per person, a data graph connects related signals (past addresses, devices, purchase behavior, life events) so a platform can resolve identity and build audiences from any starting point. Exact Match's graph covers 250M+ U.S. consumer profiles.
Household Graph
A household graph groups individual consumer profiles that live at the same address into a single family or household unit. It is used to model shared purchasing behavior, avoid duplicate outreach to the same home, and build audiences around household-level attributes like family size or homeownership rather than a single individual.
Device Graph
A device graph links multiple devices (a phone, a laptop, a tablet) back to the same individual or household. It allows a platform to recognize that a visitor on mobile and a visitor on desktop are the same person, which supports cross-device targeting, frequency capping, and more accurate audience counts.
Identity Graph
An identity graph is the umbrella structure that connects a person's identifiers (emails, phone numbers, devices, cookies, physical addresses) into one persistent profile over time. It is the foundation identity resolution runs on: the richer and more current the identity graph, the more accurately a platform can match, enrich, and target real people instead of anonymous signals.
Data Quality & Enrichment
Data Enrichment
Data enrichment is the process of appending missing information (a phone number, a physical address, a job title, or a behavioral attribute) to an existing customer or lead record. It takes a partial record and fills in the gaps using a third-party data graph, turning a name and email into a complete profile a sales or marketing team can actually act on.
Data Append
A data append adds specific missing fields to an existing list or database by matching each record against a larger reference dataset. Common append types include email append, phone append, and demographic append. The output is the original list with new columns filled in, rather than a brand-new dataset.
Email Append
Email append matches an existing customer or lead list (typically containing a name and mailing address, or a name and phone number) against a data graph to fill in a missing, verified email address for each record. It is commonly used to reactivate an offline customer list for email marketing.
Phone Append
Phone append matches an existing list against a data graph to add a verified phone number to records that only have a name, address, or email. It is used to enable outbound calling, SMS marketing, or telemarketing campaigns from a list that was originally collected without phone data.
List Enrichment
List enrichment is the broader practice of taking a partial customer or prospect list and appending any missing data, contact fields, demographic attributes, or behavioral signals, so the list becomes usable for a specific marketing or sales motion. It typically combines several append types (email, phone, demographic) in a single pass.
Data Decay
Data decay is the natural loss of accuracy in a database over time as people change jobs, emails, phone numbers, and addresses. Every contact database decays continuously: without regular refreshes, an increasing share of records go stale, bounce, or point to the wrong person. Refresh frequency is one of the most important factors in evaluating any data provider.
Data Hygiene
Data hygiene is the ongoing practice of keeping a database accurate: removing duplicate records, correcting formatting errors, verifying contact information, and suppressing records that should no longer be contacted (such as deceased individuals). Clean data improves deliverability, campaign accuracy, and the reliability of any analysis built on top of it.
Deduplication
Deduplication identifies and merges duplicate records within a database that represent the same person or household, even when the records were entered inconsistently (different name spellings, an old address, a nickname). Removing duplicates prevents redundant outreach, keeps reporting accurate, and reduces wasted spend on contacting the same person twice.
Reverse Append
A reverse append starts from a single known identifier, like an email address or phone number, and appends the remaining profile: name, mailing address, and demographic attributes. It is the mirror image of a standard email or phone append, which instead start from a name and address to find contact information.
Data Types & Sources
PII (Personally Identifiable Information)
PII, or personally identifiable information, is any data that can identify a specific individual, such as a name, email address, phone number, physical address, or government ID number. Handling PII responsibly, including honoring opt-out requests and complying with regulations like CCPA, is a baseline requirement for any consumer data platform.
Hashed Email
A hashed email is an email address that has been run through a one-way cryptographic function (commonly SHA-256) so it can be matched against other hashed records without exposing the original, readable email address. Hashed identifiers are widely used for privacy-safe audience matching between platforms, such as uploading a customer list to an ad network without sharing plain-text emails.
First-Party Data
First-party data is information a business collects directly from its own customers or website visitors: purchase history, form submissions, on-site behavior, and account details. Because it comes from an owned relationship, first-party data is generally the most accurate and privacy-durable data a company has, and it is often the starting point for enrichment or lookalike audience building.
Third-Party Data
Third-party data is collected by an entity that has no direct relationship with the individual, then aggregated and made available to other businesses. A consumer data graph like Exact Match's is a third-party data source: it supplements a company's first-party data with additional identifiers, behaviors, and attributes gathered from a broader ecosystem of sources.
Zero-Party Data
Zero-party data is information a customer deliberately and proactively shares with a business, such as answers to a preference quiz, a stated purchase intention, or profile details entered voluntarily. It differs from first-party data (observed behavior) because the customer chose to disclose it directly, which makes it especially reliable for personalization.
Firmographic Data
Firmographic data describes attributes of a company rather than an individual: industry, employee count, revenue range, and location. It is the business equivalent of demographic data and is commonly used in B2B targeting, though consumer-focused platforms typically prioritize individual-level behavioral and demographic data instead.
Audience & Targeting
Audience Segmentation
Audience segmentation is the practice of dividing a broader population into smaller groups that share common characteristics, such as demographics, behaviors, or purchase intent, so marketing can be tailored to each group rather than treated as one-size-fits-all. Effective segmentation improves message relevance and campaign performance across every channel.
Lookalike Audience
A lookalike audience is a new group of people who share behavioral or demographic similarities with an existing customer list, without necessarily having any direct relationship with the business yet. Platforms build lookalike audiences by analyzing the traits of current best customers and finding other individuals in a data graph who exhibit those same traits.
In-Market Audience
An in-market audience is a group of people currently showing behavioral signals associated with active buying intent for a specific product or category, as opposed to a broad demographic audience that may or may not be ready to buy. In-market audiences are typically built from daily-refreshed behavioral data rather than static demographic filters.
Behavioral Targeting
Behavioral targeting reaches people based on what they actually do (browsing patterns, purchase history, content engagement) rather than only who they are demographically. It groups individuals into behavioral clusters that reflect real-world actions, which tends to predict future purchase behavior more reliably than demographic targeting alone.
Psychographic Data
Psychographic data describes a person's interests, values, lifestyle, and attitudes rather than their demographic facts (age, income, location) or their raw behavior (page views, purchases). It is used to understand why an audience segment behaves the way it does, which helps refine messaging and creative beyond simple demographic targeting.
Intent Data
Intent data captures behavioral signals that indicate a person is actively researching or preparing to make a purchase decision, such as visiting comparison pages, searching for related terms, or engaging with relevant content. It lets marketing and sales teams prioritize outreach to people showing real buying signals instead of contacting a cold, undifferentiated list.
Purchase Intent Data
Purchase intent data is a specific category of intent data focused on signals tied directly to an imminent buying decision, rather than general research or awareness-stage behavior. It is typically the highest-value, most time-sensitive segment of an in-market audience, since the underlying behavior suggests a purchase is likely to happen soon.
Cross-Entity Audience
A cross-entity audience combines signals across multiple related entities, such as a household, a company, and its individual employees, to build a more complete targeting profile than any single entity alone would provide. It is useful when purchase decisions are influenced by more than one person or dimension.
Modeling & Prediction
Bayesian Behavioral Modeling
Bayesian behavioral modeling starts from an existing customer list, identifies the behavioral profile shared by the best-fit customers, then continuously updates the probability that other people in a broader data graph currently match that same profile. Unlike a static lookalike model, it produces a real-time, in-market list that refreshes daily as new behavioral signals arrive, rather than a one-time snapshot.
Predictive Lead Scoring
Predictive lead scoring ranks leads by the likelihood they will convert, using behavioral and demographic signals rather than manual, rules-based criteria alone. It helps sales and marketing teams prioritize the leads most worth pursuing first, particularly useful for smaller teams that cannot follow up with every lead individually.
Lead Scoring
Lead scoring assigns a numeric value to each lead based on attributes and behaviors that correlate with a higher likelihood of converting, such as job title, engagement level, or firmographic fit. Traditional lead scoring uses manually assigned point values; predictive lead scoring replaces or supplements those rules with a model trained on real outcome data.
In-Market Buyer Prediction
In-market buyer prediction identifies individuals who are likely to make a purchase in a specific category in the near term, based on behavioral modeling rather than a stated intent form. It is the output of applying predictive modeling, such as Bayesian behavioral modeling, to a broad consumer data graph to surface a list of high-probability buyers before they've raised their hand.
Website & Visitor Identification
Website Visitor Identification
Website visitor identification matches anonymous website traffic to real, named individuals using a pixel or script installed on the site. Instead of only seeing anonymous session data, a business can see which known individuals visited, what they viewed, and follow up directly, closing the gap left by low website form-fill rates.
Anonymous Visitor Identification
Anonymous visitor identification is the specific process of resolving a website visitor who has not filled out any form or logged in to a verified identity, using signals like IP address, device data, and a consumer data graph. A meaningful share, roughly 25 to 40 percent, of verified human visitors can typically be identified this way, depending on the site and traffic source.
Privacy, Compliance & Delivery
CCPA (California Consumer Privacy Act)
The CCPA is a California law that gives consumers the right to know what personal data is collected about them, request its deletion, and opt out of its sale. Any platform that collects or processes California consumer data, including third-party data providers, must offer a compliant opt-out mechanism and honor deletion requests within statutory deadlines.
CPRA (California Privacy Rights Act)
The CPRA expands on the CCPA with additional consumer rights, including the right to correct inaccurate personal data and to limit the use of sensitive personal information. It also created the California Privacy Protection Agency to enforce compliance. Together, CCPA and CPRA form the primary consumer privacy framework U.S. data platforms must comply with.
Data Broker
A data broker is a company that collects personal information about consumers from a variety of sources and makes it available to other businesses, without having a direct relationship with the individuals themselves. Data brokers operating in states like California and Texas are subject to registration and disclosure requirements under consumer privacy law.
Deceased Suppression
Deceased suppression removes records belonging to individuals who are known to have died from a marketing or outreach list, using data sources like the Social Security Death Master File. Suppressing deceased individuals is both a data hygiene best practice and, in many cases, a compliance requirement, since continuing to market to a deceased person's contact information can be harmful and wasteful.
Pay-As-You-Go Data
Pay-as-you-go data pricing charges for consumer data access based on a flat recurring fee rather than requiring a long-term annual contract or per-seat licensing. It gives smaller teams and agencies access to enterprise-grade data without the upfront commitment that traditional data providers typically require.
Data Onboarding
Data onboarding is the process of matching an offline dataset, such as a CRM export or a customer list, to online identifiers so it can be activated in a digital advertising or marketing platform. It is what turns a static spreadsheet of customers into an audience that can actually be targeted across email, ads, or other channels.
Branded Exports
Branded exports let a business deliver enriched or audience data to its own end clients under its own brand, rather than a third-party platform's branding. They are common among agencies that resell data services to their own customers and need the output to look native to their business.
Want the Deeper Dive?
Every term above has its own full definition page, and links out to a deeper guide or product page where one exists. For more on how identity resolution, enrichment, and audience targeting fit together, browse the full blog or see how Exact Match compares to other data providers.
See the Data Behind Every Term
One plan, everything included: every product, every feature. Get started on Unlimited for $999/mo, or $6,999/yr on annual billing.