MSc BAIS · Exam Theory Guide ← Dashboard
Exam Preparation · Spring 2026

Theory & Concept
Reference Guide

A structured overview of the frameworks, models, and arguments from your four course readings — and how they connect to your synopsis on Novo Nordisk and OpenAI.

Four Models of Sharing Economy Platforms

MIS Quarterly Executive

This article provides a 2×2 typology of sharing economy platforms, classified along two dimensions: level of control the platform exerts over participants, and intensity of rivalry it fosters among supply-side participants.

"Sharing economy platforms are not actually revolutionary. They are innovative only in that they use old mechanisms in new and 'platformed' ways."

The 2×2 Framework

Model Control Rivalry Example Value Prop
Franchiser Tight High Uber Low cost & efficiency
Chaperone Loose High Airbnb Service differentiation
Principal Tight Low Handy Low cost & risk mitigation
Gardener Loose Low Couchsurfing Community & self-organisation

Organizational Coordination Mechanisms

Mintzberg's six mechanisms: mutual adjustment, direct supervision, standardization of work processes, outputs, skills, and norms. Platforms apply these beyond traditional firm boundaries.

Market Coordination Mechanism

The price system — either dynamic (supply/demand-based) producing high rivalry, or cost-compensatory (stable) producing low rivalry. This drives the rivalry dimension.

Boundary Fluidity

The strategic asset enabling platforms to coordinate outside formal organizations. Dissolves distinctions between employee/contractor, producer/consumer, and insourcing/outsourcing.

Multi-Sided Platforms (MSPs)

Platforms serving two or more distinct user groups where each side creates value for the other. Network effects are central: more users on one side attracts more on the other.

Key Concept: Boundary Fluidity

Traditional organizational boundaries were rigid — strictly separating employees from contractors, insourcing from outsourcing, products from services. Digital platforms exploit the loosening of these distinctions. This fluidity is what allows Uber to treat drivers like quasi-employees (via standardization of outputs) while classifying them legally as independent contractors.

Five Lessons for Incumbents:

  • Understand the strategic intent of sharing economy competitors
  • Complement product portfolios with services (servitization)
  • Access new modes of innovating via open platforms
  • Engage consumers in value creation
  • Ensure strategic fit by optimizing coordination mechanisms

Rethinking the Firm: The Digital Operating Model

Competing in the Age of AI — Chapter 2

This chapter argues that digital firms (Ant Financial, Ocado, Peloton) represent a fundamentally new template for value creation and delivery — one where software, data, and AI replace labor as the primary operational bottleneck, enabling near-zero marginal cost scaling.

Business Model vs. Operating Model

Business Model = How a firm creates and captures value (the theory / strategy).
Operating Model = How a firm delivers that value (the practice / execution).

The value of a firm is constrained by its operating model. Traditional firms hit bottlenecks at scale, scope, and learning — the three operating challenges. Digital firms remove these bottlenecks by replacing human labor on the critical path with software and algorithms.

Scale

Delivering value to as many customers as possible at lowest cost. Digital operating models achieve near-zero marginal cost per additional user — the critical advantage over traditional firms.

Scope

The range of activities and services a firm offers. AI platforms achieve economies of scope by repurposing data across multiple product lines simultaneously (e.g., Ant Financial → payments → loans → insurance).

Learning

Continuous improvement through data feedback loops. AI systems improve with every interaction, enabling constant innovation without proportional increases in headcount.

Removing the Human Bottleneck

When algorithms replace labor on the critical path, the firm decouples growth from headcount. Ant Financial's 3-1-0 loan system: 3-minute application, 1-second approval, zero human interaction.

The Three Case Examples

  • Ant Financial: AI-driven financial services at massive scale. Data platform aggregates behavioral data across billions of transactions to power fraud detection, credit scoring, and loan qualification. Demonstrates extreme scale and scope.
  • Ocado: Online grocery delivery using AI, robotics, and algorithmic routing. Demonstrates how AI can remove human bottlenecks from a physically intensive business. Their algorithm runs thousands of routing calculations per second.
  • Peloton: Combines physical hardware with digital content and community networks. Demonstrates scope expansion and learning through data — fitness classes → nutrition → health services.
"In a digital operating model, the employees do not deliver the product or service; instead, they design and oversee a software-automated, algorithm-driven digital 'organization' that actually delivers the goods."

Algorithmic Labor & Information Asymmetries

International Journal of Communication

Through a nine-month empirical study of Uber drivers, this paper reveals how algorithmic management, combined with systematic information asymmetries, creates structures of soft control over workers — while the platform publicly frames the relationship as one of entrepreneurship and freedom.

Information Asymmetry as Power

Uber holds far more information than its drivers — about surge pricing accuracy, fare calculations, deactivation thresholds, and hourly guarantee eligibility criteria. Drivers receive selective, often opaque data through the app. This asymmetry is structural and intentional: it enables Uber to coordinate worker behavior without assuming the legal obligations of an employer.

Key manifestations:

  • Blind passenger acceptance: Drivers accept rides before seeing destination or fare, taking on financial risk without information.
  • Surge pricing opacity: Drivers cannot distinguish between real-time demand signals and predictive (often inaccurate) demand forecasts.
  • Unilateral rate changes: Uber changes fares and commissions without driver negotiation or consent.

Algorithmic Management

The term coined by Lee et al. (2015) and extended here: management functions are automated and delegated to software systems rather than human managers. This includes:

  • Nudging: Push notifications citing "high demand" to prevent drivers logging off — framed as information, functioning as instruction.
  • Rating systems: Passengers act as de facto middle managers, with their ratings directly affecting driver deactivation eligibility. Drivers need 4.6/5 to remain active.
  • Gamification: Surge pricing maps create lottery-like emotional engagement, driving continued platform participation.
  • Performance metrics: Weekly reports comparing individual drivers to "top drivers" on acceptance rate, hours online, fares/hour.

Soft Control

Indirect behavioral shaping through design choices, information filtering, and algorithmic nudges — without formal managerial directives. Workers feel they have freedom but operate within highly constrained choice architectures.

Emotional Labor

Drivers suppress their own emotional responses to perform hospitality for passengers in exchange for high ratings — substituting for tips that Uber's design discourages. Service work managed through peer surveillance.

The Entrepreneurship Rhetoric

Uber frames drivers as "entrepreneur-partners" with "freedom and flexibility." Rosenblat & Stark show this rhetoric obscures a managed workforce, deflecting regulatory scrutiny and employment law obligations.

Platform Disintermediation

The platform replaces traditional middle management with automated systems, while displacing accountability from the firm to individual workers. Workers bear risk but lack negotiating power.



Likely Exam Questions

Click any question to reveal a hint on which theories to draw from and how to structure your answer.

Sharing Economy · Theory Application

Using Constantiou et al.'s four models, classify Uber as a platform. How does its specific combination of coordination mechanisms enable its competitive advantage, and what ethical tensions does this create?

Draw on: Franchiser model (tight control + high rivalry), standardization of outputs as primary mechanism, dynamic pricing as market mechanism. Connect to Rosenblat & Stark for the ethical layer — boundary fluidity allows Uber to claim driver independence while exerting quasi-employment control. Structure: Define the model → explain each dimension → link to value proposition → introduce tension via information asymmetry.
▾ Show hint
Digital Firm · Operating Model

Iansiti & Lakhani argue that digital firms remove the "human bottleneck" from value delivery. What does this mean, and what are the implications for traditional firms in established industries?

Draw on: Scale, scope, and learning dimensions of the operating model. The marginal cost argument (near-zero cost to serve additional users). Ant Financial's 3-1-0 system as an extreme example. Use Ocado as an example of a "physical" industry being digitized. Implications: workforce restructuring, new governance needs, competitive pressure on incumbents who cannot achieve same scalability.
▾ Show hint
Algorithmic Labor · Power Dynamics

Rosenblat & Stark describe Uber's management of drivers as exhibiting "soft control." What mechanisms produce this, and why is the concept of information asymmetry central to understanding gig economy labor relations?

Draw on: The specific mechanisms — blind passenger acceptance, surge pricing opacity, rating systems as delegated managerial surveillance, nudging via push notifications. Define soft control (Boltanski & Chiapello). Information asymmetry: Uber holds data that drivers need to make informed decisions, but withholds or obscures it. Connect to the entrepreneurship rhetoric as ideological cover.
▾ Show hint
Synopsis · Integration Question

Apply the concept of the "AI factory" to Novo Nordisk's partnership with OpenAI. What are the strategic opportunities and what risks does your analysis identify?

Draw on: Iansiti & Lakhani's four components of the AI factory. Your case: data pipeline (clinical trials, genomics, manufacturing metrics), algorithm development, experimentation platform, software infrastructure. Opportunities: scalability of drug discovery, removing human bottlenecks from routine analysis. Risks from your synopsis: algorithmic bias in underrepresented patient populations, data sovereignty with a third-party AI provider, asset specificity creating dependency on OpenAI.
▾ Show hint
Strategic Management · Comparative

Both Uber and Novo Nordisk use algorithmic systems to coordinate work. Using the course readings, compare how algorithms shape the relationship between the firm and its workers/partners differently in each case.

Draw on: Rosenblat & Stark for Uber — algorithmic management of a precarious, disaggregated workforce using information asymmetry. Iansiti & Lakhani for Novo Nordisk — algorithms on the critical path of product delivery, with humans moving into design and oversight roles. Key difference: in Uber, workers are kept in the dark to maintain soft control; in Novo Nordisk's envisioned model, upskilling means workers must understand and collaborate with AI (Stelmaszak's codependence). Also bring in boundary fluidity — both involve dissolving traditional organizational distinctions.
▾ Show hint
Platform Economics · Network Effects

What is the role of network effects in sharing economy platforms? How do direct and indirect network effects differ, and why are they strategically central to platforms like Airbnb and Uber?

Draw on: The MSP (multi-sided platform) section from Constantiou et al. Direct network effects: same-side (more drivers → more value for drivers competing for riders). Indirect/cross-side: more drivers → more value for passengers, more passengers → more value for drivers. The self-reinforcing loop. Connect to competitive moats — once a platform achieves critical mass, the network effects create enormous switching costs. Link also to Iansiti & Lakhani's N*E*M formula (users × engagement × monetization).
▾ Show hint
Ethics & Governance

What ethical challenges arise when algorithms are used to manage human labor or influence clinical decisions? How should organizations govern these systems?

Draw on: Rosenblat & Stark for labor: power asymmetry, lack of accountability, emotional labor externalization, deactivation without due process. Your synopsis for clinical: algorithmic bias from unrepresentative training data, data security with third-party providers, accountability when AI de-prioritizes a drug compound that would have succeeded. Governance responses: transparency in algorithmic decision-making, worker/patient representation in system design, regulatory frameworks, multidisciplinary oversight.
▾ Show hint
MSc Business Administration & Information Systems Copenhagen Business School · 2026