Session 06 · Using Smart Data for Digital Strategy

How AI and Connected Products Are Reshaping Firms

A structured guide to the core theories, frameworks, and exam questions from Porter & Heppelmann (2015), Iansiti & Lakhani (2021), and Medaglia's lecture.

Key Question 1 What are the economic implications of implementing AI in the firm?
Key Question 2 How does competition change with connected products?
01

Core Concepts at a Glance

The key building blocks for session 6
Iansiti & Lakhani

The AI Factory

A scalable decision engine that treats decision-making as an industrial process. Analytics convert data into predictions, which guide or automate operational actions — moving humans off the critical path.

Porter & Heppelmann

Smart Connected Products

Products with physical, smart, and connectivity components. They can monitor, control, optimise, and act autonomously — and they fundamentally reshape the value chain of any firm that makes them.

Iansiti & Lakhani

Datafication

Systematically extracting data from activities that already occur in any business. The Nest thermostat datafied home climate. Netflix datafied TV watching. Ride-sharing datafied urban transport.

Medaglia

AI Collision

When an AI-driven firm serves the same customers as a traditional firm with a similar (or better) value proposition but a far more scalable operating model — e.g. Amazon vs retailers, Ant Financial vs banks.

Iansiti & Lakhani

Virtuous Cycle of AI

More usage → more data → better algorithms → better service → more usage. This self-reinforcing loop creates formidable scale and learning advantages that are hard for incumbents to replicate.

Porter & Heppelmann

Technology Stack

The new infrastructure smart products require: product hardware/software, connectivity, the product cloud (with analytics engine, app platform, database), plus security and integration with enterprise systems.

02

The AI Factory

Iansiti & Lakhani (2021), Chapter 3 — the scalable decision engine of the digital firm

"The AI factory treats decision-making as a science. Analytics systematically convert internal and external data into predictions, insights, and choices."

— Iansiti & Lakhani, Competing in the Age of AI
Fig 1 The virtuous cycle — how the AI factory self-reinforces
AI Factory virtuous cycle A circular diagram showing: more data leads to better algorithms, better algorithms lead to better service, better service leads to more usage, more usage leads to more data. AI Factory virtuous cycle More data from every interaction Better algorithms predictions improve Better service value to users More usage platform grows

The four essential components that make the AI factory work:

  1. Data pipeline

    Gathers, inputs, cleans, integrates, processes, and safeguards data in a systematic, sustainable, and scalable way. Netflix, for example, ingests billions of ratings, stream plays, queues, and search terms. The challenge: data from incumbent firms is often siloed, fragmented, and inconsistent.

  2. Algorithm development

    Algorithms generate predictions about future states or actions of the business. Three families: supervised learning (predict from labelled data — e.g. spam filters, churn prediction), unsupervised learning (find hidden groupings — e.g. customer segmentation), reinforcement learning (explore/exploit trade-off — e.g. Netflix artwork personalisation, AlphaGo).

  3. Experimentation platform

    Hypotheses about improvements are tested as randomised controlled trials (A/B tests) to distinguish genuine causal effects from spurious correlation. Google runs over 100,000 experiments per year. Netflix tests every significant product change before release.

  4. Software infrastructure

    Embeds the pipeline in a consistent, componentised software and computing environment and connects it to internal and external users via APIs. A well-designed data platform lets an agile team build a new application in days rather than years.

Exam questions — AI Factory

1
What distinguishes an AI factory from traditional data analytics? What makes it "scalable"?
Hint: think about the industrial analogy — mass production of decisions, continuous feedback loops, and the elimination of human bottlenecks on the critical path.
2
Explain the difference between supervised, unsupervised, and reinforcement learning using one business example for each.
Hint: supervised = churn prediction / loan default; unsupervised = customer segments / anomaly detection; reinforcement = Netflix artwork selection / dynamic pricing.
3
Why is the experimentation platform described as a "necessary component" of the AI factory? What goes wrong without it?
Hint: correlation ≠ causation. Without A/B testing, an algorithm that correlates with a target metric might just be capturing a pre-existing pattern, not driving change.
4
Describe "datafication" with an example. Why do incumbent businesses often struggle with this first step?
Hint: data exists but is fragmented across silos, lacks common identifiers, has inconsistent formats. Executives consistently underestimate the cleaning and integration investment.
03

Smart Connected Products

Porter & Heppelmann (2015) — what they are and what they enable
Fig 2 Three layers of a smart connected product
Three layers of smart connected products Stacked diagram showing: physical components at the base, smart components in the middle, connectivity components at the top. Arrows show how each layer amplifies the layer below. Connectivity components Ports, antennae, protocols, networks — enables cloud & product communication Smart components Sensors, microprocessors, data storage, controls, embedded OS, UI Physical components Mechanical and electrical parts — the traditional product amplifies Enables: Monitor Control Optimise Autonomy

The four new product capabilities these layers unlock:

① Monitor

Products report on their own condition and environment — generating insights previously unavailable (real-time engine health, soil moisture, usage patterns).

② Control

Users can operate products remotely, customise function and performance, and operate in hazardous or hard-to-reach environments.

③ Optimise

Algorithms substantially improve performance, utilisation, and uptime — and how a product works with related products in a broader system (smart building, smart farm).

④ Autonomy

Products learn, adapt to users and environment, service themselves, and operate independently — the combination of all three prior capabilities.

Exam questions — Smart Connected Products

1
Describe the three core components of a smart connected product and explain how connectivity "amplifies" the smart layer.
Hint: smart components exist within a product. Connectivity extends them outward to the cloud and back — unlocking remote control, continuous software upgrades, and system-level optimisation.
2
Porter & Heppelmann say data from smart products is "valuable by itself, yet its value increases exponentially when integrated with other data." Illustrate this with a concrete example.
Hint: farm humidity sensor data + weather forecast + commodity prices → optimal irrigation; vehicle location + service needs + parts inventory → predictive maintenance scheduling.
3
What is "evergreen design" and how does it differ from traditional product development? What competitive advantage does it create?
Hint: traditional products are fixed between generations. Smart products continuously improve via software. Tesla's autopilot and suspension updates are the canonical example — no recall required.
04

Transforming the Value Chain

How smart connected products reshape every function in the manufacturing firm
Fig 3 Value chain transformation — from function to function
Function Before (traditional product) After (smart connected product)
Product development Mechanical engineering; fixed-generation design; physical testing only Interdisciplinary systems engineering; evergreen design; continuous real-world quality monitoring; low-cost software variability
Manufacturing Discrete process ending at shipment; standardised physical assembly Continuous process — the cloud stack is part of the product; smart factories with networked machines; software loaded post-shipment
Logistics RFID/barcode scanning; periodic location tracking Continuous real-time tracking of location, condition, and environment; drone delivery; optimised fleet routing
Marketing & Sales One-time transaction focus; segment by demographics; customer surveys Continuous value delivery; fine-grained usage-based segmentation; product-as-a-service models; focus on systems not products
After-sale service Reactive, on-site visits; technician diagnoses on arrival; expensive second trips Remote & preventive service; one-stop visits; AR-assisted repair; predictive analytics (Diebold ATMs); new service revenue streams
HR Mechanical engineers; traditional sales and service skills Software engineers, data scientists, UX designers; new clock speeds; new compensation norms; co-location with tech hubs

New organisational units that emerge

New structure

Unified Data Organisation

Led by a Chief Data Officer. Consolidates data collection, aggregation, and analytics across all functions. No single function can manage the volume, complexity, and security of data by itself. (Gartner: 25% of large firms by 2017.)

New structure

Dev-Ops

Bridges product development ("dev") with IT and manufacturing operations ("ops"). Manages continuous product operation, software updates, patches, and new post-sale services. Shortens release cycles, runs cloud infrastructure.

New structure

Customer Success Management

Monitors product use and performance to ensure customers extract maximum value — especially critical in product-as-a-service models to drive renewals. Uses the product's data stream rather than waiting for complaints.

Exam questions — Value chain & organisation

1
Porter & Heppelmann argue that "manufacturing becomes a permanent process." What do they mean, and why?
Hint: smart products can't function without the cloud technology stack. The manufacturer must operate and improve that stack for the product's entire life — it never ends at shipment.
2
What is a "Customer Success Management" unit? Why can't traditional sales or service units perform this function?
Hint: traditional sales is incentivised on the transaction, not ongoing value. Traditional service reacts to failures. CSM proactively monitors usage data to prevent churn and expand value — different incentives and tools.
3
What is a "data lake" and why are conventional approaches like spreadsheets and database tables inadequate for smart product data?
Hint: smart product data is unstructured and heterogeneous — sensor readings, temperatures, GPS coordinates, video, warranty history. A data lake stores data in native formats; conventional tools require structured schemas up front.
05

How Incumbents Should Respond

Becoming an AI company — Microsoft, Fidelity, and the five principles
Fig 4 Four stages of operating model transformation (Iansiti & Lakhani)
Four stages of digital operating model transformation Four boxes from left to right: Stage 1 Siloed data, Stage 2 Pilot, Stage 3 Data hub, Stage 4 AI factory. An upward-sloping arrow underneath indicates increasing performance impact. Transformation stages are: Improvement, Demonstration, Optimization, Transformation. Stage 1 Siloed data scattered Excel files Stage 2 Pilots analytics demos Stage 3 Data hub integrated platform Stage 4 AI factory automated decisions Low Performance Improvement Demonstration Optimization Transformation

Iansiti & Lakhani's research shows AI maturity leaders had a 55% gross margin vs 37% for laggards, and 11% net income vs 7%. The five principles for transformation:

  1. One strategy

    Strategic clarity and full commitment — no plan B. Not about spinning off an AI division; it requires rebuilding the core. Nadella's "The cloud is our future and we have fundamentally no choice" is the model.

  2. Architectural clarity

    Everyone must understand what the future operating architecture looks like. Data must be integrated, not fragmented. Without a consistent data catalog and clear standards, AI is impossible to safeguard and scale.

  3. Agile, product-focused organisation

    AI-centric applications must embed a deep understanding of their settings — like any product. Agile teams building on a shared data platform can deploy new applications in days. Gone are the years-long custom IT builds.

  4. Capability foundations

    Hire data scientists, software engineers, and — crucially — data and analytics product managers who can bridge business context and technical execution. The market for this talent is hot and competitive.

  5. Clear, multidisciplinary governance

    AI's societal impacts (privacy, bias, fairness, cybersecurity) must be governed across legal, engineering, and product functions together. Microsoft's six AI principles (fairness, reliability, privacy, inclusiveness, transparency, accountability) are the benchmark.

"For traditional firms, becoming a software-based, AI-driven company is about becoming a different kind of organisation — one accustomed to ongoing transformation. This is not about spinning off a new organisation, setting up the occasional skunkworks, or creating an AI department."

— Iansiti & Lakhani, Chapter 5

Exam questions — Incumbents & transformation

1
What is an "AI collision"? Use an example to explain why the incumbent is structurally disadvantaged.
Hint: Ant Financial vs banks — same service (lending), but Ant has a scalable AI operating model with no branches, no loan officers, and real-time data. The incumbent has fixed costs and siloed legacy data that it can't easily replicate away.
2
Why does Medaglia suggest that becoming a platform is one viable path for incumbents? What does this achieve?
Hint: a platform lets the incumbent interconnect products/services, generate data repositories, use the cloud for storage, and begin building the data pipeline that an AI factory requires. It breaks down product silos.
3
Compare the AI factory's implications for the firm's operating model with what Porter & Heppelmann say about smart connected products' implications for the value chain. What is common?
Hint: both require: centralised data, new cross-functional coordination (IT&R&D, or dev-ops), new roles (CDO, data scientist, customer success), shift from reactive to proactive/predictive, and continuous rather than episodic operations.
4
Microsoft tripled its stock price in Nadella's first three years as CEO. What specific operational changes drove this, and which of the five transformation principles does each reflect?
Hint: embracing open source (architectural clarity + one strategy), moving Azure to the core (one strategy), building supply chain capability for data centres (capability foundations), Core Services Engineering / DelBene's role (agile product-focused org), and AI governance / Brad Smith (multidisciplinary governance).
06

Synthesis & Cross-Concept Questions

Questions that require linking frameworks — the hardest and most likely oral exam territory

High-difficulty synthesis questions

S1
A traditional car manufacturer says: "We have sensors in our cars. We're already a smart connected products company." Critically assess this claim using Porter & Heppelmann's framework and Iansiti & Lakhani's AI factory concept.
Hint: having sensors ≠ having a data pipeline, algorithm development capability, or experimentation platform. The manufacturer may have the physical layer but lack the cloud stack, the data organisation, the dev-ops unit, and the analytics feedback loop. Tesla is the comparison.
S2
Porter & Heppelmann and Iansiti & Lakhani both argue that data is the decisive asset of the 21st century firm. But they emphasise different challenges in generating and using it. Compare their perspectives.
Hint: P&H focus on the challenge of integrating product data with enterprise data (the data lake, the new technology stack, the unified data organisation). I&L focus on the pipeline challenge (cleaning, normalisation, bias), the algorithm development process, and experimentation. Both see siloed data as the key barrier for incumbents.
S3
The Gartner Hype Cycle (referenced in Medaglia's slides) suggests technologies are often over-hyped then under-valued before reaching productive use. How does this concept apply to AI in enterprises today?
Hint: relate to the four-stage transformation model (siloed → pilot → hub → factory). Many firms are at the "trough of disillusionment" after failed AI pilots (stage 2 failures) because they haven't addressed the underlying data architecture required for stage 3+.
S4
Porter & Heppelmann's 10 strategic decisions for smart connected products include: "Should the company pursue an open or closed system?" How does this decision interact with Iansiti & Lakhani's argument about network effects and data learning loops?
Hint: open systems attract more partners and users → more data → better AI → more competitive moat (I&L's logic). But open systems also risk sharing the data advantage with competitors (P&H's concern). SmartThings (open) vs Apple HomeKit (more closed) is a live example.

Session readings

  • 📄
    Porter & Heppelmann (2015) — How Smart, Connected Products Are Transforming Companies Harvard Business Review, October 2015. Focus: value chain transformation, technology stack, new organisational structures.
  • 📘
    Iansiti & Lakhani (2021) — Chapter 3: The AI Factory Competing in the Age of AI. Focus: data pipeline, algorithm development (supervised / unsupervised / reinforcement), experimentation platform, infrastructure.
  • 📘
    Iansiti & Lakhani (2021) — Chapter 5: Becoming an AI Company Competing in the Age of AI. Focus: Microsoft's transformation, five principles, four stages, AI maturity index, Fidelity case.
  • 🎓
    Medaglia — Using Smart Data for Digital Strategies (Session 6 slides) CBS lecture. Focus: AI/ML definitions, AI factory, datafication, AI collision, smart connected products, incumbents' response.