Case Study 11.1: Acxiom and the Database Nation

The Company You've Never Heard Of That Knows Everything About You

Background

In 2002, journalist Robert O'Harrow Jr. published a landmark investigation in The Washington Post that introduced many Americans to a company they had never heard of: Acxiom Corporation. Based in Conway, Arkansas — an unlikely address for a company at the center of the information economy — Acxiom had been quietly building one of the most comprehensive databases of personal information ever assembled. O'Harrow described it as knowing more about the American population than any institution except the federal government. By 2012, Acxiom held profiles on approximately 300 million Americans, with an average of 1,500 data points per person. By 2023, its databases covered more than 700 million people worldwide.

Acxiom was not a household name, and it was not supposed to be. Its business was not selling products to consumers; its business was selling consumers — or, more precisely, data about consumers — to other businesses. Its clients included retailers, banks, insurance companies, political campaigns, government agencies, and healthcare organizations. What they purchased was intelligence: information that helped them target, assess, price, and communicate with specific individuals or population segments.

How Acxiom Builds Its Profiles

Acxiom's data collection operation functions through aggregation — the combination of information from thousands of different sources into unified consumer profiles. The company has never detailed all of its data sources publicly, but regulatory filings, academic research, and investigative journalism have established the following major categories:

Retailer transaction data: Acxiom purchases transaction records from retailers and catalog companies — what you bought, how much you paid, how often you bought. Loyalty card programs, which offer discounts in exchange for behavioral tracking, are a major source. A customer who joins a pharmacy loyalty program has, in the fine print of the membership agreement, typically authorized the pharmacy to share or sell their purchase data with third parties.

Financial records: Through partnerships with financial service providers, Acxiom obtains data about credit card categories (types of merchants where cards are used), loan types, and payment behaviors. The specific transactions are usually not sold — this would violate banking regulations — but behavioral categories derived from them are.

Public records: Property records, voter registration files, vehicle registration records, court records, and business license filings are public documents in most U.S. jurisdictions. Acxiom employs teams to collect, digitize, and integrate this information systematically.

Surveys and self-reported data: Acxiom runs its own surveys and purchases survey data from companies whose forms — warranty registrations, sweepstakes entries, magazine subscription forms — contain detailed lifestyle questions.

Third-party data purchases: Acxiom purchases data from other data brokers, magazine subscription companies, catalog companies, and businesses that have accumulated their own customer databases and are willing to sell access.

The combination of these sources produces profiles that typically include: name, age, address history, household composition, home ownership status, estimated home value, vehicle information, income bracket, education level, occupation category, interests and hobbies (modeled), political affiliation (modeled), religious affiliation (modeled), health interests (modeled), estimated net worth, purchase history categories, and dozens of proprietary scoring variables.

The "AboutTheData" Experiment

In 2012, facing growing public concern and regulatory pressure, Acxiom took an unusual step: it launched a consumer-facing portal called "AboutTheData.com" that allowed individuals to see some of the data the company held about them. The portal represented, the company said, a commitment to transparency and consumer control.

In practice, the experience was more complicated. Users who accessed their profiles found that the data was organized into broad categories rather than specific records — "Possibly Present," "Likely," or "Very Likely" for various characteristics. Some users found their profiles surprisingly accurate; others found significant errors. A 2012 investigation by the Associated Press found users whose income brackets were wrong, whose household composition was incorrect, and whose interest categories reflected a life they didn't recognize.

The portal also illustrated the limitations of self-policed "transparency." Acxiom showed users some of their data — but not all of it. The most sensitive inferred categories, the data sold to specific clients, and the proprietary scoring variables were not visible. Users could see enough to understand that a profile existed but not enough to understand how it was being used or what decisions it was influencing.

The opt-out mechanism was also notable. Users who wanted to remove their data from Acxiom's marketing databases could request an opt-out — but this opt-out applied only to Acxiom's marketing services division, not to its other business lines. Data shared with government agencies, law enforcement, financial institutions, or healthcare companies was governed by different agreements and different opt-out mechanisms, if any existed at all.

The Law Enforcement Dimension

One of Acxiom's most significant and least-publicized data relationships is with government and law enforcement agencies. After September 11, 2001, Acxiom became a significant contractor for federal agencies seeking commercial data to augment their intelligence capabilities. The company provided data to the Transportation Security Administration to support its CAPPS II passenger screening program and was later investigated by the Senate Judiciary Committee for providing data to the FBI without adequate legal authorization.

This law enforcement relationship exemplifies function creep in the clearest possible terms. The data that Acxiom collects for marketing purposes — to help retailers sell products more effectively — is the same data that government agencies use for security screening, fraud investigation, and law enforcement purposes. The consumer whose loyalty card purchases informed a coupon targeting algorithm also, simultaneously, informed a federal passenger risk scoring system. They had agreed to neither use.

The Government Accountability Office's 2013 report on data brokers found that federal agencies were spending hundreds of millions of dollars annually to purchase commercial data — from Acxiom and others — to supplement their own databases. This practice creates what legal scholar Paul Schwartz called a "commercial surveillance subsidy" for government intelligence: the government effectively outsources data collection to commercial actors who operate with fewer legal restrictions, then purchases the results.

The Pricing Structure and What It Reveals

Acxiom's commercial pricing illustrates the different values placed on different types of individuals. The company's data products are priced not just by volume but by the predicted value of the individuals described. A list of high-net-worth individuals with confirmed interest in luxury products commands significantly higher prices than a list of general population records. A list of people who have recently been diagnosed with specific health conditions (available from health data specialists who partner with Acxiom) can sell for many times the price of demographic data.

This pricing structure reflects — and reinforces — the commercial logic of social sorting. Individuals are valued differently based on their predicted commercial utility. The data economy does not treat all people equally; it treats people according to their estimated value as consumers. This has downstream consequences: high-value consumer profiles receive premium marketing, preferential offers, and targeted incentives. Low-value profiles receive different marketing — or are systematically ignored by algorithms that deprioritize them as not worth the acquisition cost.

Analysis Questions

1. Acxiom describes itself as a "data and technology company" that helps businesses "connect with the right customers." Evaluate this framing. What does it emphasize? What does it obscure? How might a data subject describe the same company's activities?

2. The "AboutTheData" portal represented what Acxiom called a commitment to transparency. Using the concepts from Chapter 11, evaluate whether this transparency was substantive or performative. What would genuine transparency look like?

3. The fact that Acxiom's data is used simultaneously for commercial marketing and law enforcement screening is an example of function creep. Who, if anyone, consented to the law enforcement use of this data? What does the absence of meaningful consent reveal about the structure of the data broker system?

4. Acxiom's pricing structure assigns higher value to some consumers than to others. What are the ethical implications of a system in which human individuals are commercially valued differently based on behavioral and demographic profiles? How does this relate to the concept of social sorting introduced in Chapter 5?

5. The chapter argues that the data economy requires structural rather than individual responses. Does Acxiom's existence support this argument? Could individual consumer choices — refusing loyalty cards, avoiding Acxiom's client companies — meaningfully constrain Acxiom's operation? Why or why not?

Connections

This case study connects directly to: - The data pipeline model (Section 11.7) - Function creep (Chapter 5) - Social sorting (Chapter 5) - Government use of commercial surveillance data (Chapter 22) - Legal frameworks for data brokers (Chapter 28)

Case Study 11.1 | Chapter 11 | Part 3: Commercial Surveillance