Topic 14 · Deep Dive

14.1 What Makes Personal Data Valuable

Fragments Are Cheap; Joined Records Are Not

Every visit, install, and purchase emits fragments: a third-party cookie set by an embedded script, a one-pixel image request whose URL query encodes your ID and the page you were on, a server-to-server conversion postback that names you by hashed email, or a fingerprint assembled from canvas rendering, fonts, and user-agent entropy. No single fragment is valuable. What gets harvested is a stable identifier that any site loading the same tag can reuse to recognize you again.

Inside the Aggregation Pipeline

Brokers run continuous ETL pipelines. Inputs arrive as delimited files: loyalty and warranty lists, app event streams, postal change-of-address feeds, ad server logs, and occasionally breach dumps. The pipeline normalizes fields, then performs identity resolution: deterministic and probabilistic matching over emails, phone numbers, name-plus-DOB pairs, and device IDs, often comparing salted hashes rather than plaintext, to fuse duplicate rows into a single person node on an identity graph. Enrichment then attaches categories -- income band, home value, new parent, recently relocated -- each scored by confidence. The output is one keyed record, resold many times.

Two Buyers, One Record

On the regulated side, data marketplaces and demand-side platforms buy those categories for targeting: lookalike modeling, life-event campaigns, suppression lists. On the fraud side, the same fused rows are repackaged as fullz -- name, date of birth, partial SSN, address history, phone and utility records -- the raw material for account takeover, SIM-swap pretexting, and synthetic identity applications. A breached password plus a broker-supplied address history is all a takeover artist needs to survive a security-question review.

The Economics That Keep the Pipes Running

An isolated phone number trades for fractions of a cent; a verified, multi-source record enriched with wealth and life-event categories is worth orders of magnitude more, and fresh-breach fullz price higher still. That spread is the entire incentive structure: brokers must ingest widely and resolve identities aggressively because value concentrates in fusion, not in any single source. Your defensive surface therefore sits upstream, at trackers and permissions, not at the database itself.

Architecture Diagram

Personal data value chain user activity visits, apps, purchases trackers cookies, pixels, fingerprints broker DB identity resolution + fusion targeting ID-theft market raw fields are cheap; fused verified records are not
Value is created at the fusion step: the same resolved row feeds both ad targeting and the fraud market.

Key Takeaways

« Back to Topic 1414.2 »