Analysis Report Three - Privacy and Customer Profiling

Author

Nora Miller

Executive Summary

In today’s digital economy, efforts to make money from customer data have led to a basic conflict between company profits and consumer privacy. Many retailers are now becoming lucrative advertising platforms (Gabel, Simester, and Timoshenko 2024), but this intense focus on customer profiling can cause consumer concerns and draw attention from regulators (Morey, Forbath, and Schoop 2015). This report reviews recent literature on data strategies to show that trust isn’t built through surface-level corporate messaging (Ward 2024). Instead, transparent data practices are needed. The report wraps up with practical examples of customer profiling using the Northwind database, covering table joins, data aggregation, and group filtering.

Introduction

The readings show that companies are changing how they see their customers. Gabel, Simester, and Timoshenko (2024) explain that big retailers like Amazon and Walmart are using huge amounts of transaction data to become “retail media” platforms, turning what people do into ad money (Gabel, Simester, and Timoshenko 2024). At the same time, major tech companies are creating the AI “friend economy” by building chatbots that grab users’ attention and gather their data by reflecting their opinions back to them (Higgins 2025).

But collecting all this data brings real worries from customers. Morey, Forbath, and Schoop (2015) point out that since people know they’re being watched online, companies should stop hiding what they do and focus on being open about their practices (Morey, Forbath, and Schoop 2015). Just claiming to be ethical doesn’t convince anyone. Ward (2024) finds that using too many “trust words” like integrity and responsibility in corporate reports can actually backfire and attract scrutiny from regulators (Ward 2024). Building real trust means offering clear, fair value—not just making promises.

Privacy vs Profiling

Trying to create detailed customer profiles often clashes with today’s standards for privacy. If companies don’t get this balance right, they can end up facing serious legal and financial trouble.

A recent real-world example of this is the regulatory action taken against unauthorized profiling. In 2022, Sephora had to pay a $1.2 million fine under the California Consumer Privacy Act (CCPA) for not telling customers it was selling their data to third-party advertisers to create profiles, and for not honoring customers’ requests to opt out of tracking (State of California Department of Justice 2022).

This real-world case backs up what the Harvard Business Review says: companies can’t just hope for forgiveness anymore—they need permission up front (Morey, Forbath, and Schoop 2015). Today, profiling needs to be open and clear. Even Apple has stepped in with App Tracking Transparency (ATT), which makes apps directly ask users before tracking them across different apps (Apple Inc. 2021). The whole industry is moving away from secret data collection toward a model where people have to give clear, informed consent.

Profiling Example

To demonstrate how companies actually build these profiles, we will analyze the purchasing behavior of a specific high-volume Northwind customer: Ernst Handel (CustomerID: ERNSH).

#| label: Myquery1

SELECT 
    c.CategoryName,
    SUM(od.Quantity) AS TotalItemsPurchased,
    SUM(od.Quantity * od.UnitPrice) As TotalSpent
FROM Orders o
JOIN [Order Details] od ON o.OrderID = od.OrderID
JOIN Products p ON od.ProductID = p.ProductID
JOIN Categories c ON p.CategoryID = c.CategoryID
WHERE o.CustomerID = 'ERNSH'
GROUP BY c.CategoryName
HAVING SUM(od.Quantity * od.UnitPrice) >1000
ggplot(data = myquery1, aes(x = reorder(x = CategoryName, TotalSpent), y = TotalSpent)) +
  geom_col(color = "darkgreen", alpha = 0.8) + coord_flip() +
  labs(
    title = "Ernst Handel: Individual Transaction Values Over Time",
    x = "Product Category",
    y = "Total Amount Spent ($)",
    subtitle = "Aggregated using GROUP BY and SUM"
  ) +
  theme_minimal()

Recommendations for Industry

Based on the literature and our analysis, organizations must implement the following strategic adjustments:

  1. Show, Don’t Tell: Skip the generic “trust words” like ‘ethical’ or ‘integrity’ in materials for customers or regulators, since those can actually raise suspicions (Ward 2024).

  2. Implement Fair Value Exchange: If you’re building a retail media network, make it clear what users get in return for their data—like lower prices or better recommendations (Gabel, Simester, and Timoshenko 2024).

  3. Respect Opt-Outs and Avoid Forced Consent: As the Sephora CCPA case shows, companies need to respect privacy requests instead of hiding them in hard-to-find settings. Profiling should always be based on real consent (State of California Department of Justice 2022).

References

Apple Inc. 2021. “App Tracking Transparency.” https://developer.apple.com/app-store/app-privacy-details/.
Gabel, Sebastian, Duncan Simester, and Artem Timoshenko. 2024. “How Retailers Became Ad Platforms.” Harvard Business Review.
Higgins, Tim. 2025. “In a Lonely World, Elon Musk, Mark Zuckerberg and Even Microsoft Are Vying for Affection in the New ’Friend Economy’.” The Wall Street Journal.
Morey, Timothy, Theodore Forbath, and Allison Schoop. 2015. “Customer Data: Designing for Transparency and Trust.” Harvard Business Review 93 (5): 97–103.
State of California Department of Justice. 2022. “Attorney General Bonta Announces Settlement with Sephora as Part of Ongoing Enforcement of California Consumer Privacy Act.” https://oag.ca.gov/news/press-releases/attorney-general-bonta-announces-settlement-sephora-part-ongoing-enforcement.
Ward, Lisa. 2024. “Companies Using ’Trust Words’ Are about 15% More Likely to Receive a Comment Letter from the SEC.” The Wall Street Journal.