Business Research Proposal: User Perceptions of Surveillance Capitalism

Recommended Frameworks: Kano Model & Conjoint Analysis

• Framework Teaching:
  • Kano Model: Think of this as a framework for sorting features based on how they impact user satisfaction. Instead of just asking if users "like" something, it classifies attributes into five categories:
    1. Must-be: Expected features that cause dissatisfaction if absent (e.g., a messaging app must be able to send messages).
    2. Performance: The more you have, the better (e.g., more cloud storage space).
    3. Attractive: Unexpected delights that create satisfaction but don't cause dissatisfaction if absent (e.g., a fun, new photo filter).
    4. Indifferent: Users don't care about these one way or the other.
    5. Reverse: Features that users actively dislike. Their presence causes dissatisfaction.
  • Conjoint Analysis: This is a market research technique that reveals what customers truly value by asking them to make trade-offs. Instead of asking a direct question like, "Is privacy important?" (to which most people will say yes), it presents realistic scenarios and forces a choice. For example, "Would you prefer a service with hyper-personalized content that tracks your browsing history, or a service with generic content that does not?" By analyzing thousands of these choices, you can mathematically calculate the precise value users place on each attribute, such as privacy.
• Applicability Explanation:
  • The Kano Model is perfectly suited to identify which specific data collection practices cause the most user concern. By framing data practices as "features" of a service, you can use the model to find which ones fall into the Reverse category—those that users actively dislike. This directly answers your question about which aspects are most concerning.
  • Conjoint Analysis is the ideal tool to test the perceived value of "free" services versus privacy. It quantifies the trade-off, allowing you to understand exactly how much convenience or functionality users are willing to exchange for a given level of data privacy. It moves beyond simple opinion polling to reveal underlying preferences.
• Key Information to Collect:
  • A comprehensive list of data collection and usage practices employed by major tech companies.
  • User reactions to the presence or absence of these data practices (for the Kano Model).
  • User choices between different hypothetical service packages, each with varying levels of data privacy and service features (for Conjoint Analysis).
  • Demographic data (age, country of residence, education, etc.) to segment the results.

Part 1: Information Collection

Web Search

• Search Queries:

  • "common data collection methods by social media and tech giants"

  • "how tech companies use personal data for profit"

  • "examples of surveillance capitalism business models"

  • "studies on user attitudes towards data privacy 2024"

  • "summary of GDPR and CCPA data subject rights"

• Purpose of Queries:
  • These searches will help you compile a list of 15-20 specific and clearly worded data practices (e.g., "Using your private messages to train AI models," "Selling your location history to data brokers," "Tracking your browsing activity across other websites"). This list is the backbone of your entire study.
  • Understanding existing privacy regulations (like GDPR) and previous studies provides context for your findings.

User Survey

• Interview Subjects:
  • A statistically significant and demographically diverse sample of users of popular "free" digital services (e.g., Google, Meta, TikTok). Ensure representation across different age groups, locations, and self-reported levels of technical expertise.
• Interview Purpose:
  • To gather quantitative data on user preferences and concerns regarding the data practices identified in your web search.
• Core Survey Questions:
  1. For the Kano Analysis (to identify concerns): For each data practice on your list, you will ask a functional/dysfunctional pair of questions.
     • Example Practice: "This service shares your activity data with third-party partners for advertising purposes."
     • Question 1 (Functional): "How would you feel if the service DID this?"
       • Answers: (a) I like it, (b) I expect it, (c) I am neutral, (d) I can tolerate it, (e) I dislike it.
     • Question 2 (Dysfunctional): "How would you feel if the service DID NOT do this?"
       • Answers: (a) I like it, (b) I expect it, (c) I am neutral, (d) I can tolerate it, (e) I dislike it.
     • Analysis Purpose: The combination of answers for each practice will allow you to categorize it as Must-be, Performance, Attractive, Indifferent, or Reverse. A high number of "Reverse" classifications signals a major user concern.
  2. For the Conjoint Analysis (to measure trade-offs): This section will present users with a series of choices. You will define key attributes with different levels.
     • Example Attributes & Levels:
       • Personalization Level: (a) Fully personalized feed, (b) Partially personalized, (c) Chronological/Generic feed.
       • Location Tracking: (a) Always on, (b) Only while using the app, (c) Never tracked.
       • Data Sharing with 3rd Parties: (a) Yes, for advertising, (b) No, never.
     • Sample Question: "Which of the following two services would you prefer?"
       • Service A: Fully personalized feed, Location never tracked, Data shared with 3rd parties.
       • Service B: Chronological feed, Location tracked only while using, Data never shared.
     • Analysis Purpose: By analyzing how users choose across 10-15 of these scenarios, you can calculate the numerical value (known as "utility") they assign to each level. This will explicitly show how much "Personalization" they are willing to sacrifice for "No Data Sharing."

Part 2: Information Analysis

1. Analyze Kano Results to Rank User Concerns:
   • For each data practice, tabulate the paired answers from all respondents using a standard Kano evaluation table. This will give you the percentage of users who classify that practice as Reverse, Indifferent, Must-be, etc.
   • Create a ranked list of the data practices from highest percentage of "Reverse" classifications to lowest. This is your deliverable for "which aspects of surveillance capitalism cause most user concern."
2. Analyze Conjoint Results to Quantify Value Trade-offs:
   • Use a statistical software package (many online survey tools have this built-in, or you can use specialized software) to run the conjoint analysis on the choice data.
   • The output will provide "utility scores" for each attribute level. A higher utility score means it is more valued. For example, you might find the utility for "Data never shared" is +1.5, while the utility for "Fully personalized feed" is +0.8.
   • This allows you to create a clear hierarchy of what users truly value, quantifying the privacy vs. convenience trade-off and addressing the "perceived value of 'free' services" part of your question.
3. Synthesize Findings into the Final Report:
   • The Perception Gap: Compare the list of common industry practices you found during your web search with your ranked list of user concerns from the Kano analysis. The overlap—practices that are both common and highly disliked—represents the primary gap between corporate practice and user desire.
   • Demographic Segmentation: Cross-tabulate the results from both your Kano and Conjoint analyses with the demographic data you collected. This will allow you to generate insights such as: "Users aged 18-24 are more tolerant of location tracking than users over 50," or "Users in Europe place a higher value on data minimization than users in North America."
   • Develop Strategic Recommendations: Based on your findings, formulate actionable advice. For example:
     • For Policymakers: Highlight the top 3-5 most disliked data practices as priority areas for regulatory review.
     • For Tech Companies: Identify data practices that are highly disliked by users but have low utility in the conjoint model. Suggesting that companies abandon these "low-value, high-concern" practices could be an easy way to build user trust.