Identity Data Glossary

The process of connecting fragmented data signals — emails, phone numbers, MAIDs, IP addresses, postal addresses — to a single unified person or household record. Identity resolution uses deterministic matching (exact identifier overlap), probabilistic bridging (statistical inference across shared attributes), or a combination of both to link records across sources and devices. Best-in-class providers score every resolved link with a confidence value so downstream teams can set quality thresholds. Learn more at BIGDBM Identity Resolution.

A database that links every known identifier for a person — email addresses, phone numbers, physical addresses, device IDs, IP addresses — into a connected record with confidence-scored edges. An identity graph is the underlying infrastructure for resolution. It records not just the identifiers but the strength of the connections between them, so downstream consumers can set quality thresholds. BIGDBM's graph spans 200M+ identities and 5B+ verified connections. Learn more at BIGDBM Trusted Identity Graph.

A resettable, device-level identifier assigned by mobile operating systems: AAID (Android Advertising ID) on Android and IDFA (Identifier for Advertisers) on iOS. MAIDs allow advertisers to target or frequency-cap users across mobile apps without using personally identifiable information. When linked to an identity graph, a MAID can be resolved to a named individual, household, or email address. BIGDBM holds billions of MAID linkages, refreshed weekly.

The Android operating system's resettable advertising identifier — one of the two primary MAID types alongside Apple's IDFA. Assigned by Google Play Services, the AAID enables cross-app tracking and audience targeting on Android devices. Users can reset or opt out of personalized advertising using their AAID at any time in device settings. BIGDBM resolves AAIDs to household-level PII, property records, and behavioral intent signals.

Apple's device-level advertising identifier for iOS devices. Since iOS 14.5 (App Tracking Transparency), apps must request explicit user permission before accessing the IDFA, significantly reducing availability. When available, the IDFA enables deterministic cross-app targeting and measurement on Apple devices. Because ATT opt-in rates average 25–40%, IDFAs require supplemental probabilistic bridging to achieve meaningful coverage at scale.

A one-way cryptographic transformation of an email address using MD5, SHA-256, or SHA-1. Hashed emails allow identity matching across platforms without exposing raw PII — two parties can compare MD5(email) to confirm a shared identity without exchanging the underlying address. BIGDBM resolves hashed emails at 80%+ match rate against a database of 700M+ email records covering all three major hash formats. The 80%+ figure reflects clean, current email data; older or role-based lists will match lower.

The percentage of submitted records that a resolution provider successfully links to a known identity in its graph. Match rate = matched records ÷ total submitted records × 100. A match rate of 60–70% is typical for average-quality consumer lists; BIGDBM achieves 80%+ on hashed email and higher on clean, current consumer PII. Match rate is heavily influenced by data recency, formatting consistency, and identifier type. See How to Audit Your Match Rates for a step-by-step improvement guide.

BIGDBM's four-dimensional confidence scoring framework applied to every identity record and link. RFIS stands for Recency (how recently the signal was observed), Frequency (how often it has appeared), Intensity (how strong it was), and Strength (how many independent sources confirmed it). Each dimension is scored 0–100 and combined into a composite that allows teams to set thresholds before activation — eliminating the black-box resolution problem. Learn more at RFIS Scoring Model.

An identity resolution method that links records based on exact matches of shared identifiers — two records with the same email address, phone number, or SSN fragment are connected with high certainty. Deterministic matching produces fewer false positives than probabilistic methods but achieves lower coverage because exact matches require clean, current, consistently formatted data. BIGDBM applies deterministic resolution first in every match sequence before falling back to probabilistic bridging for expanded reach.

An identity resolution method that infers connections between records based on shared attributes — same device cluster, overlapping IP history, similar name + location + age — without requiring exact identifier overlap. Probabilistic matching expands coverage significantly but introduces false positive risk that must be managed with confidence scoring. BIGDBM uses probabilistic bridging as a secondary layer after deterministic resolution, and every probabilistically inferred link carries an RFIS confidence score.

BIGDBM's phone scoring system that evaluates whether a number is active, mobile vs. landline, DNC-registered at the federal and state level, and associated with a real identified individual. PQL scoring allows outreach teams to filter phone lists before dialing — reducing waste, protecting caller reputation, and managing TCPA compliance risk. Scores are segmented into tiers so teams can balance reach against risk for their specific use case. Available through the Intelligence Marketplace.

BIGDBM's email scoring system that evaluates deliverability, inbox placement probability, recency of engagement, and bounce and spam classification risk for each email address. EQL scoring enables senders to suppress low-quality records before deployment — protecting sender domain reputation and improving measured open and click rates. Like PQL, EQL scores are tiered so minimum quality thresholds can be set independently per campaign or channel.

Behavioral signals collected from a consumer's or business's online activity — content consumption, search queries, product comparisons, review reading — that indicate in-market interest in a category or product right now. Unlike demographic data (who someone is), intent data captures what someone is actively researching. BIGDBM's intent data is IAB-classified, daily-refreshed, and identity-resolved to named individuals rather than anonymous devices or probabilistic household clusters. Learn more at What is Intent Data?

The process of appending additional attributes to an existing record using a third-party data source. A CRM record with only a name and email might be enriched with verified phone number, physical address, household income estimate, property ownership status, device identifiers, and intent signals — all returned in a single API call. BIGDBM's Real-Time Enrichment APIs return enriched fields in 100–300ms per record.

Data collected directly by a business from its own customers and prospects — web behavior, purchase history, email engagement, survey responses, account profiles. First-party data is typically the most accurate and consent-clear data a business owns, but it covers only people who have already interacted with the brand, limiting its use for prospecting new audiences. Enriching first-party data with verified third-party appends expands its depth without sacrificing the consent baseline. See First-Party vs Third-Party Data.

Data collected by a company with no direct relationship with the individual — compiled from public records, web crawls, purchase behavior aggregators, and consent-based opt-in networks, then licensed to other businesses. Quality varies enormously by provider. High-quality third-party data is sourced with consent management, CCPA/CPRA compliance, and continuous freshness validation built into the infrastructure. BIGDBM sources from 80+ underlying providers and applies RFIS confidence scoring to every record before delivery.

California's primary consumer privacy law, effective January 1, 2020, granting California residents the right to know what personal information is collected about them, the right to request deletion, and the right to opt out of the sale of their personal information. CCPA applies to businesses meeting size or revenue thresholds. Data providers supplying personal information to third parties must maintain opt-out processing and communicate deletion rights downstream. See CCPA Compliance Checklist.

California's expansion of CCPA, effective January 1, 2023, that added the right to correct inaccurate personal information, strengthened restrictions on sensitive personal information (SPI including precise geolocation, health data, racial origin, and sexual orientation), created the California Privacy Protection Agency (CPPA), and introduced purpose limitation rules. BIGDBM's data operations infrastructure was built to comply with both CCPA and CPRA from the ground up — not retrofitted after the fact.

A software application that centralizes customer data from multiple sources — CRM, web analytics, email, POS — into unified customer profiles that marketing and product teams can activate against. Unlike identity graphs (which focus on cross-source resolution infrastructure), CDPs focus on activation workflows and segmentation. Many CDPs integrate with identity resolution providers to improve their underlying match quality. See Identity Graph vs. CDP for a detailed comparison.

The process of matching an offline customer file — CRM records, loyalty program members, purchase histories — to online identifiers (device IDs, cookies, MAIDs) so that offline-known audiences can be targeted in digital channels. Data onboarding is typically performed by identity resolution providers and enables deterministic digital targeting without relying on probabilistic device graphs. Turnaround for file-based onboarding at BIGDBM is 24–48 hours with field-level RFIS transparency on every matched record.