Session 04 1.5 hours Analysis

Beyond the Five Forces

Ecosystems, business models and the game your rivals play

Every firm sits in a web of players who create value with it and then fight it for a share: customers, suppliers, rivals, and the firms whose products make yours worth more. This session maps that web, asks how a firm can change the game being played in it, and asks what a strategist should do when no forecast can say how the game will end.

Session 3 rated the US auto industry with Porter’s five forces and ended on a loose thread: complements. Rothaermel counts them as a sixth force; Porter says they work through the five. Then Musk said Tesla was “open” to other carmakers using its Superchargers if they shared the cost (Sull & Reavis, 2019, p. 14). The moment a rival can also be a customer of your complement, five boxes stop being enough. Today you draw the whole game instead.

Start with question 1 Readings & preparation Learning outcomes & visual lab

The three questions

  1. 01Who else shapes the value Tesla can capture?
  2. 02How should a firm play, and change, the game its rivals play?
  3. 03What should a strategist do when the future cannot be forecast?

Each question belongs to one reading. Brandenburger and Nalebuff answer the first two: their value net maps who shapes the value, and PARTS shows how to change the game. Courtney, Kirkland and Viguerie answer the third. The Tesla case is where all three get argued out loud.

Start here · Session 4

Learning outcomes

By the end of this session, students should be able to:

  1. Construct a value net and explain when the same organisation can play more than one role. Draw and check the value net.
  2. Distinguish value created from added value, and identify a concrete PARTS lever for changing the game. Classify a game-changing move.
  3. Use best responses or backward reasoning to assess a rival’s likely action, with explicit assumptions. Work the strategic game.
  4. Diagnose residual uncertainty and match a strategic posture and action portfolio to it. Distinguish the four uncertainty levels.
  5. Defend a Tesla ecosystem recommendation using case evidence, a plausible response and a measurable learning trigger. Build a posture-and-move argument.

Start the existing session · Try the visual practice lab · Prepare with the assigned readings

The activities help you practise these outcomes. Browser feedback supports learning; formal assessment follows the course outline and instructor guidance.

Recall → connect → carry forward · Sessions 1–3

The strategy story so far: Sessions 1–3

We have moved from choosing a coherent position, to defining its purpose and ambition, to testing the industry around it. Those steps explain why Nila might enter and where profit could be squeezed; they also expose relationships that a static threat assessment cannot settle. Nila Foods remains a fictional Indian meal-service, and its next decision is whether cooperation can strengthen its focused position.

One continuing classroom story. Nila Foods is a fictional Indian meal-service for working adults. Follow its choices from Session 1; later expansion ideas are proposals to test.

  1. Session 1 · What is strategy?

    Operational effectiveness improves similar activities; strategy requires positioning, trade-offs and fit. State objective, scope and advantage, then compare intended plans with emergent action using evidence.

    Nila Foods: Nila proposes preordered weekday meals on clustered routes, declines instant citywide delivery and studies actual customer behaviour.

  2. Session 2 · Framework for Strategic Analysis

    Purpose and values guide what endures, while a quantified ambition opens a same-date gap from the baseline; closing that gap requires external and internal evidence.

    Nila Foods: Nila separated easing meal burdens and honest food-safety disclosure from its growth target, then compared that target with its projected baseline.

  3. Session 3 · Industry Analysis: The Fundamentals

    Industry boundaries, PESTEL and five forces connect external change to price, cost and investment; complements and possible responses then raise questions about the wider game.

    Nila Foods: Nila traced ingredient-price pressure, supplier bargaining, lunch substitutes and easy entry before asking whether a neighbouring seller could also help it.

Why this leads to Session 4

A neighbouring meal seller might compete for Nila’s subscribers while sharing a collection point that makes both services easier to use. Session 3 tells us to trace how that arrangement changes the forces; it does not settle who joins, what rules they accept, or how either firm responds. Session 4 therefore maps the value net, explores PARTS and matches commitments to what remains uncertain.

Retrieve it before moving on

How can sharing a collection point with a rival fit Nila’s earlier strategy, and what remains to be analysed?

Compare your reasoning with a model answer

The arrangement fits only if it reinforces Nila’s dependable, focused service and respects its values and growth logic (Sessions 1–2). Five forces help trace effects on costs, rivalry and customer options (Session 3). We still need to identify each player’s roles and added value, negotiate rules and consider responses under uncertain demand.

Question 01 · Brandenburger & Nalebuff (1995)

Who else shapes the value Tesla can capture?

Brandenburger and Nalebuff begin with a complaint about language. Business borrows its words from war and sport, and both mislead: business is not about winning and losing. Firms can do spectacularly well without anyone else failing, and they can fail however well they play if they are playing the wrong game (1995, p. 57). The game, in their words, “is all about value: creating it and capturing it” (p. 59).

To see who shapes that value, they draw a map they call the value net. On the vertical axis are the firm’s customers and suppliers: resources flow up from suppliers, products flow down to customers, and money flows the other way. On the horizontal axis are players the firm interacts with but does not transact with: substitutors and complementors (p. 59).

The second idea is added value. Take the value created when everyone is in the game, remove one player, and see how much value the rest can still create. The difference is that player’s added value, and in a game without fixed rules no player can take away more than its added value (p. 58). Profit, in other words, follows added value, not value.

“Successful business strategy is about actively shaping the game you play, not just playing the game you find.”
Brandenburger & Nalebuff (1995), p. 58

Their method has three parts: a mind-set (coopetition, looking for win–win as well as win–lose moves), a map (the value net) and a method (PARTS) (p. 70). Question 1 uses the map; question 2 uses the method.

CUSTOMERS

Buy the firm’s products. With suppliers they form the vertical axis, where firms cooperate to create the pie and compete to divide it (p. 60).

SUPPLIERS

Provide labour, materials and other resources. Customers and suppliers are symmetric: any strategy towards one has a counterpart towards the other (p. 71).

SUBSTITUTORS

Alternatives from whom customers may buy, or to whom suppliers may sell. Coca-Cola and Pepsi are substitutors for drinkers; Coca-Cola and Tyson Foods are substitutors for carbon dioxide, which one uses to carbonate and the other to freeze chickens (p. 60).

COMPLEMENTORS

Players whose products make customers value yours more, or make suppliers keener to supply you. Faster chips and more powerful software raise each other’s value. American and United are complementors to Boeing: each one’s order helps the other’s plane get built (p. 60).

Why new words

Call a player a “competitor” and you think only about fighting it. “Substitutor” names the relationship without that bias, and complementors, its natural counterpart, are the players traditional analysis most often overlooks (p. 60). One player can hold several roles at once: American and United compete for passengers and complement each other when they buy planes.

The half most managers miss

Managers see substitutors only as enemies and complementors only as friends. Both pictures are half right. A substitutor can be a partner, as GM’s credit card showed Ford, and a complementor can be squeezed, as 3DO squeezed the hardware makers who built its game console (pp. 58, 60–61, 63). Question 2 returns to both stories.

Where Session 3 stopped, and why the value net starts there

Session 3 · five (six) forces

Complements raise value. Who keeps it?

Rothaermel counts complements as a sixth force; Porter says they act through the five. Both agree a complement can grow demand. Neither says how the extra value is split between Tesla and the firms whose products make a Tesla worth more. Added value is the tool for that split.

Session 3 · the hand-off

A rival that could also be a customer

If other carmakers paid to use Tesla’s Superchargers (Sull & Reavis, 2019, p. 14), they would be substitutors for car buyers and customers of the charging network at the same time. A five-forces diagram can only draw them as rivalry. The value net can draw both roles.

Grant (2018) · the critique

Structure without moves

The central criticism of the five forces, Grant writes, is that they ignore competitive interaction: rivalry becomes a variable linking structure to profit, not a sequence of decisions by firms that anticipate each other (p. 91). And industry factors explain less than 20% of the variation in firms’ return on assets (p. 84). Industry analysis still matters; it is not the whole picture.

Value is not added value: Nintendo and the Game Boy

Cars, televisions and video recorders create far more value in the world than Game Boys. Yet between July 1990 and June 1991 Nintendo’s average stock-market value, 2.4 trillion yen, was higher than Sony’s (2.2 trillion) or Nissan’s (2.0 trillion). Creating value is not enough; profit comes from capturing it, and Nintendo captured a giant slice of a largish pie by keeping its own added value high and every other player’s low (Brandenburger & Nalebuff, 1995, pp. 65–66).

  • Customers. Retailers such as Toys R Us and Wal-Mart were concentrated, so Nintendo did not fill every order: 33 million cartridges in 1988 against demand for 45 million.
  • Complementors. A security chip and a limited number of licences kept outside game developers plentiful and replaceable.
  • Suppliers. Old-generation chips made chip suppliers a commodity; once Mario was a hit, licensed characters mattered less.
  • Substitutors. The largest installed base, plus exclusivity from developers, left rival consoles no added value to offer (p. 65).

Activity · value net

Draw Tesla’s value net

Eight players from the two cases. Give each one every role it holds towards Tesla: customer, supplier, substitutor, complementor. Most hold more than one. The map fills in as you decide; then check each player against the evidence.

Tesla's value net, as you have drawn it Tesla in the centre; customers above, suppliers below, substitutors to the left and complementors to the right. Each box lists the players you have placed in that role.

Layout after Brandenburger & Nalebuff’s value net (1995, p. 60), redrawn for this session. Your choices, not a verdict, until you check.

0 of 8 checked

The two articles, and the textbook behind them

One question each

Brandenburger and Nalebuff answer who shapes the value and how to change the game. Courtney, Kirkland and Viguerie answer what to do when the game’s end cannot be forecast. Each tab rebuilds the author’s central argument so you can work it rather than read it. The last two tabs connect the readings and show what Grant’s textbook chapter adds.

Added value, played as a card game

The article’s central argument fits in one classroom game. Adam holds 26 black cards; each of his 26 students holds one red card. The dean pays $100 for every red card matched with a black one. Without Adam there is no game, so his added value is the whole $2,600. Without any one student, one match is lost, so each student’s added value is $100. The added values sum to $5,200, but there is only $2,600 to divide, so nobody gets all of theirs; by symmetry each side expects about half, $50 a card (Brandenburger & Nalebuff, 1995, p. 64).

Then Adam publicly burns three of his black cards. The pie shrinks to $2,300, but three students will now go unmatched, so no single student is essential and each one’s added value falls to zero. Adam can offer a 90:10 split and look generous (pp. 64–65). Burn cards below, and then let the students bargain as one.

Two principles sit under the game. In games with rules, look forward and reason backward: play out every reaction to your move, including your own reply to theirs. In games without rules, you cannot take away more than you bring. Both demand allocentrism: thinking from inside the other players’ heads, and asking what you bring to them rather than what they bring to you (p. 58).

Activity · the card game

Black cards Adam burns

The 50:50 split at no burned cards and the 90:10 split after burning are Brandenburger and Nalebuff’s; applying 90:10 to other numbers of burned cards, and an even split when the students bargain as one, are this page’s illustration.

Not just a card trick

Nintendo ran the same play on retailers, game developers and chip suppliers: it kept every other player short of cards (p. 65).

The students’ counter

Adam’s 90% depends on the students not getting together. If they did, that would change the players, and it would be an excellent strategy for them (pp. 64–65).

Reading task

Pick one participant in Tesla’s ecosystem. Ask what changes if it disappears, then propose one invitation, contract change, information release or change of boundary, and say who would respond defensively. A proposal to test, not a prediction about the case.

Ecosystems · Grant (2018), ch. 4 · Further

Complements, ecosystems and who captures the value

Grant starts from the asymmetry the value net builds in. A substitute lowers the value of a product; a complement raises it. A printer is useless without ink cartridges, a car without fuel. Once you accept that, the strategic question is no longer whether complements matter but how the value they create is shared between the firms that supply the pieces (Grant, 2018, p. 86). The value net names the players; this section asks which pieces are worth owning.

01 · Complements

The split decides the profit

In 1990s video games, independent developers supplied most of the software and most of the consumer value, yet Nintendo took most of the system’s profit because it controlled the operating system and the making of cartridges. In personal computers the opposite happened: IBM’s open architecture let Windows become the proprietary standard, and the PC drifted towards a commodity (p. 86).

Grant’s rule: keep your own product monopolised, differentiated and in short supply, and encourage competition, commoditisation and excess capacity among your complements (p. 87). Apple’s ownership of iOS lets it choose which apps reach the iPhone and take 30% of their revenue; Nestlé lost control of Nespresso capsules once its patents stopped protecting them.

02 · Ecosystems

Value migrates; look for the bottleneck

A business ecosystem is the community of organisations, institutions and individuals that affect the enterprise, bound by codependence and always evolving. Value moves between its parts with technology, regulation and customer preferences, and firms can push it. The prize is a bottleneck: an activity that creates significant customer value and that the firm can dominate (Grant, 2018, p. 88, drawing on Jacobides).

  • Be the guardian of quality (Intel Inside).
  • Be irreplaceable; keep others replaceable (Apple’s “supply chain atomisation”).
  • Ride shifts in customer needs.
  • Redefine the value chain (IKEA moved assembly to the customer).

03 · Platforms, in brief

When the interface is the business

In digital systems, competition is often between platforms, the interfaces that link the parts. Complementors favour the platform with the most users and users favour the one with the most complements; that loop is the main source of winner-take-all markets (Grant, 2018, pp. 85–87).

The course’s earlier platforms deck put the contrast simply: a pipeline designs, makes and sells (Bajaj Auto, Godrej Appliances); a platform connects producers with consumers (Uber, Airbnb, Flipkart, Zomato). Two-sided platforms are Session 8’s topic, with Eisenmann, Parker and Van Alstyne (2006) and the NVIDIA case.

The value-capture test · Grant (2018), pp. 90–91

How much would the network lose if you left it?

Grant’s value-capture model, which he traces to work by Brandenburger and Stuart, puts bounds on what a firm can take from its network. The ceiling is the value the network would lose if the firm walked away: Brandenburger and Nalebuff’s added value. The floor is what the firm could add to an alternative network. Investments with competitive intent raise the value the firm adds; actions with persuasive intent change how much the others are willing to concede. Ask it of Tesla and of a battery maker: which one does the other need more?

Activity · ecosystem map

Which pieces of Tesla’s ecosystem are bottlenecks?

Eight pieces of the ecosystem around a Tesla, as the cases describe them up to about 2019. Apply Grant’s two tests to each: does it create significant value for the customer, and could one player dominate it because it is scarce or hard to substitute? A piece that passes both is a bottleneck: worth owning, or worth making sure a rival does not own. The map changes as you decide. Then compare your read with the case evidence and with what Tesla actually did.

0 of 8 marked as bottlenecks

Business models · Grant (2018); Rothaermel (2024) · Further

The business model: how a firm gets paid inside its ecosystem

A model is a simplified description of something real, so a business model is a simplified description of a business: its core logic for creating value. Grant quotes David Teece’s definition, “the manner by which the business enterprise delivers value to customers, entices customers to pay for value, and converts those payments to profit” (Grant, 2018, p. 88). He leaves open whether a business model is more than a strategy under another name, and uses it for one job: designing how a firm exploits the opportunities in its ecosystem (pp. 88–89).

Rothaermel puts it more bluntly: the business model is where strategy turns into action, and it explains how the firm intends to make money. He asks four questions of any model: why it creates value (its revenue and cost models), what activities deliver the offer, who performs them, and how the activities are linked (2024, pp. 476–477). In value-net terms, the who is the question that matters here: which activities the firm keeps, and which it leaves to customers, suppliers and complementors.

The catalogue of business-model types (razor and blades, subscription, freemium, the long tail and the rest) belongs to Session 13. This session keeps one question: where in the ecosystem does the money get captured?

The canvas · Grant (2018), p. 89

The Business Model Canvas treats the firm as four blocks. A business model is one integrated set of choices across them; mapping it lets you try alternative configurations.

Infrastructure

activities, resources, partners

Customers

segments, channels, relationships

Offer

the value proposition

Financial viability

revenues, costs, profit

Tesla against the industry recipe

Six business-model choices. Read across each row: what the industry did, what Tesla did, and what that did to the ecosystem around the car.

Swipe the table sideways to see all four columns.

ChoiceThe industry recipeTeslaWhat it did to the ecosystem
Distribution Franchised dealers; decades of dealer lobbying produced laws limiting how and where makers could sell (Serafeim & Migdal, 2024, Exhibit 2). Company-owned showrooms in malls, staff paid no commission, orders in store or online (p. 8). In 2019 it closed all but a few of its 100-plus stores; most Model 3 orders were online (p. 12). Removed the intermediary that, a 2016 study found, steered buyers away from EVs (Sull & Reavis, 2019, p. 6).
After-sales Dealers made 45–60% of their profits from service and parts (Sull & Reavis, 2019, p. 5). Many fixes sent over the internet; 76 US service centres, about half beside stores (Serafeim & Migdal, 2024, p. 8; Sull & Reavis, 2019, p. 14). Shrank a profit pool. In Tesla’s own accounts, services and other had a negative share of gross profit every year from 2017 to 2021 (Exhibit 11a).
The car after sale Fixed at purchase. Connectivity sold as packages: GM OnStar at $24.99 a month after a six-month trial (Sull & Reavis, 2019, p. 9). Software updates improve acceleration and braking after sale (Sull & Reavis, 2019, p. 14). Full self-driving and autopilot capability priced at $8,000 from 2016 (Serafeim & Migdal, 2024, p. 9). Moved value from hardware to software, the layer PwC warned carmakers could lose to become “dumb pipes on wheels” (Sull & Reavis, 2019, p. 9).
Energy Fuel is someone else’s business. Free Supercharging at the Model S launch (Serafeim & Migdal, 2024, p. 8); later $0.24 per kWh (Sull & Reavis, 2019, p. 14). SolarCity bought in 2016; Musk pitched “an electric car, a Powerwall and a solar roof” as the integrated future (Serafeim & Migdal, 2024, p. 10). Turned a missing complement into part of the product. Energy generation and storage was 5–10% of revenue in 2017–2021 (Exhibit 11c).
Regulation Emissions rules are a compliance cost. Regulatory credits were 2.4%, 5.0% and 2.7% of revenue in 2019–2021, but 14.6%, 23.8% and 10.8% of gross profit (Serafeim & Migdal, 2024, Exhibits 11a, 11c). A rule in the ecosystem became a margin source, paid by other regulated firms (Tesla, 2022). The exhibit workshop works out how much.
The car as a platform — Master Plan, Part Deux: owners could add their self-driving car to a shared fleet. One analyst put an owner’s earnings at up to $6,900 a year after paying Tesla a 10% cut (Sull & Reavis, 2019, p. 14). Would make Tesla the owner of a two-sided platform matching owners with riders, and a substitutor to ride-hailing firms. Planned, not built, in the cases; platforms are Session 8’s topic.

Serafeim & Migdal (2024) is abbreviated to page or exhibit numbers after its first mention in a row.

Read the table as a system

Each choice changes the ecosystem as well as the income statement. Direct sales took the dealer out of the value net; over-the-air fixes shrank the service profit pool the dealer lived on; Superchargers made a missing complement part of the product. That is why the choices hang together, and why Session 1’s point about fit applies: copying one piece is easy, copying the set is not.

But single pieces do get copied

Customers could collect their Tesla from a store or have it delivered home, “an approach other manufacturers and dealers quickly mimicked” (Serafeim & Migdal, 2024, p. 8). Brandenburger and Nalebuff add that imitation is not always bad news (1995, p. 59). The activity below asks which rows an incumbent could copy, and at what cost.

Activity · the copy test

Could an incumbent copy it, and what would copying cost?

Six pieces of Tesla’s business model, seen from a legacy carmaker around 2018. For each, decide what stands in the way of copying it. The obstacle is often not the rival itself but another player in the rival’s own value net.

0 of 6 checked

Question 02 · Brandenburger & Nalebuff (1995)

How should a firm play, and change, the game its rivals play?

Brandenburger and Nalebuff open with the US car industry Session 3 analysed. In the early 1990s, year-end rebates and dealer discounts were ruining its profits: when one maker cleared stock with incentives the others had to follow, and buyers learned to wait for them (1995, p. 58). In September 1992 GM and Household Bank launched a credit card that let holders put 5% of their spending towards a new GM car, up to $500 a year and $3,500 in all. There were 1.2 million accounts after a month and 8.7 million after two years.

The card won customers from Ford, a win–lose move. It also replaced GM’s other incentives, which raised the price a non-cardholder paid for a GM car. That gave Ford room to raise its prices, which let GM raise its own: a win–win (p. 58). When Ford and Volkswagen copied the card, they paid for it by cutting their other incentives, so GM’s customers faced higher prices elsewhere. Imitation helped GM (p. 59).

That is coopetition: looking for win–win as well as win–lose moves. Win–lose moves such as price cuts for share often backfire; once matched, they leave everyone where they started at lower prices (p. 59). To change a game, change one or more of its five elements, PARTS (p. 61).

Why win–win is worth looking for · p. 59

  1. It is relatively unexplored, so there are more opportunities to find.
  2. Others give up no ground, so they resist less.
  3. It does not force retaliation, so the new game lasts.
  4. Imitation of a win–win move helps you instead of hurting you.

Did GM plan all this, or get lucky with a loyalty programme? The authors cannot say. Their point is that the stakes are too high to leave to chance (p. 59).

P · PLAYERS

None is fixed, yourself included. Coke and Pepsi encouraged the Holland Sweetener Company to build an aspartame plant, then signed new long-term contracts with Monsanto just before the US NutraSweet patent expired, saving about $200 million a year (pp. 61–62). HSC’s mistake: it should have asked to be paid to play, up front.

A · ADDED VALUES

Raise yours or lower others’. TWA took 5 to 40 seats out of each plane in 1993 and went to first place in long-haul customer satisfaction (p. 64). Softsoap’s maker ordered 100 million pumps, both suppliers’ whole output, buying a 12–18 month head start (p. 66).

R · RULES

From law, custom, practicality or contract. Kiwi International promised to take no more than 10% of any market, so the majors gained more by accommodating it than by fighting it: judo economics (p. 67). Meet-the-competition clauses let carbon-dioxide producers earn more than their added value (pp. 67–68).

T · TACTICS

Moves that shape how others perceive the game: lifting the fog or thickening it. Murdoch announced a 25-cent New York Post and tested it on Staten Island, where sales doubled; the Daily News then raised its own price to 50 cents (p. 69).

S · SCOPE

Games are linked across markets and over time. Sega launched a 16-bit console; Nintendo took two years to answer, because a cheaper 16-bit market would have eaten the value of its 8-bit franchise (p. 70).

Tesla · a published sequence of moves

The 2006 master plan announced Tesla’s sequence in advance: a sports car, then an affordable car, then a cheaper one, plus zero-emission power (Serafeim & Migdal, 2024, p. 5). Session 2 read it as purpose. Read it now as a tactic: whose perception of the game was it meant to change, and was it lifting the fog or thickening it?

Five traps · pp. 70–71

Accepting the game you find. Assuming a change must come at others’ expense. Believing you must do something no one else can. Failing to see the whole game, complementors above all. And failing to think methodically, from the other players’ side. There is no end to the game of changing the game: others are changing it too.

Activity · PARTS

Which element of the game did they change?

Nine moves from the article and two from Tesla. Name the primary lever for each, then check. Good moves usually pull more than one lever; the feedback says when a second answer is defensible.

0 of 11 checked

Game theory · Grant (2018), pp. 91–95 · Further

The formal tools behind “look forward, reason backward”

A strategic situation is one of interdependence: what each player should do depends on what it expects the others to do. Game theory gives that situation a structure by asking four questions: who the players are, what options each has, what each player earns from every combination of options, and in what sequence the decisions are made (Grant, 2018, p. 91). Brandenburger and Nalebuff’s PARTS are a manager’s version of the same structure; Grant supplies the vocabulary for solving a game once it is drawn.

Sequence · simultaneous games

Choose without seeing the other’s choice

When players commit before they can observe each other, as with capacity booked a year ahead, you reason by best responses: for each move the rival might make, what is my best reply? A pair of moves from which neither player can gain by changing alone is a Nash equilibrium (Grant, 2018, p. 93). It need not be good for either player.

Sequence · sequential games

Move first, and reason backwards

When the second mover sees the first move, the first mover works backwards: for each move I could make, how will the rival reply, and which of those outcomes do I prefer? This is Brandenburger and Nalebuff’s first principle (1995, p. 58). Moving first is worth something only if the move is visible and hard to undo. A price cut that can be reversed next quarter is not a first move; it is an announcement.

The prisoners’ dilemma · p. 93

Grant’s example: Coca-Cola and Pepsi each choose a big or small advertising budget. Both would earn more with small budgets, but each fears the other will exploit its restraint, so both spend big. The equilibrium is worse for both, and any agreement to restrain is unstable because each gains by cheating. Two escapes: turn a one-off game into a repeated one, or change the payoffs through deterrence.

Look for the same structure in the cases. Charging stations were unprofitable to build, and stakeholders argued over who should pay; BMW and Nissan chose to back one shared fast-charging network, EVgo, open to every EV driver (Sull & Reavis, 2019, p. 7). Which escape is that?

Five ways to change the outcome · pp. 91–94

Cooperation
Co-opetition, Brandenburger and Nalebuff’s term: every relationship mixes the two. Coke and Pepsi still cooperate on policy for soda sales in schools.
Deterrence
Impose costs on moves you do not want. Only credible threats deter: a price war often hurts the incumbent more than the entrant; excess capacity is more convincing.
Commitment
Eliminate your own options so that others believe you. Airbus advertised the A380 heavily before the design was finished.
Changing the game
In June 2014 Tesla offered its patents to competitors, betting that a bigger EV market would outweigh any lost advantage (p. 94).
Signalling
Selective communication meant to shape how rivals read your moves. Threats are credible when backed by reputation; preannounced price changes can also invite cooperation. Compare Brandenburger and Nalebuff’s tactics.

Tesla · commitment and signal

Panasonic agreed to fund 30–40% of the $5 billion Gigafactory 1 (Serafeim & Migdal, 2024, p. 9): a visible, sunk, shared investment that is hard to walk back. The case also reports one reading of Musk’s 2018 pay plan as a signal rather than pay: “They were about signaling a credible commitment to Tesla’s purpose: to become a clean energy giant” (p. 10). Does a pay plan remove any options? A factory does.

How far to trust it · pp. 94–95

Game theory is rigorous, but it gives clear predictions only in stylised situations; with realistic detail it often yields no equilibrium or several. Grant judges it better at explaining the past than predicting the future. Its use is to structure the situation and show how to change the game. Commitment returns in Session 7, with Ghemawat and Pisano on sustaining superior performance.

Activity · payoff matrix

Incumbent: launch EVs now or wait? Tesla: cut price or hold?

Two players, two options each. Pick any cell to see whether either player would want to move away from it. Then solve the game, and change one element at a time: the order of moves, Tesla’s added value through its costs, and the rules on emissions. The payoffs are an illustrative profit index (higher is better), chosen to make the logic visible; they are not figures from the cases.

Order of moves

Tesla’s cost position

After: Tesla expected battery cells at $100 per kWh by the end of 2018, against the $145 GM paid LG Chem for the Bolt (Sull & Reavis, 2019, p. 13). A lower cost makes a price cut pay for Tesla.

Emissions rules

Penalised: governments tighten fleet emissions limits and announce bans on new combustion cars (Serafeim & Migdal, 2024, Exhibits 4 and 9), so a carmaker without EVs loses money whatever Tesla does.

Tesla: Cut price
Tesla: Hold price
Incumbent:Launch EVs now
Incumbent:Wait

Each cell: incumbent’s payoff, then Tesla’s. A ring marks the equilibrium once solved.

Pick a cell, or press Solve the game.
Try this sequence. Solve the simultaneous game before the Gigafactory. Switch to Tesla first: why does a first move help Tesla here, and why would a mere price announcement not do it? Then switch on the Gigafactory and try all three orders: that is a change in added value. Finally penalise waiting: that is a change in the rules. In the cases, GM planned 20 EV models by 2023 and Ford $11 billion for 16 EVs and 24 hybrids by 2022 (Sull & Reavis, 2019, p. 7). Which version of the game were they playing?

Competitor analysis · Porter’s framework, in Grant (2018), pp. 95–98

Competitor analysis: get inside the rival’s head

Brandenburger and Nalebuff’s demand to be allocentric, to put yourself in the shoes and even the heads of the other players (1995, p. 58), needs a method. In a concentrated industry the behaviour of a few close rivals dominates the environment. Game theory supplies the formal apparatus, but Grant argues that for most decisions a less formal, evidence-based prediction of what rivals will do is enough (2018, p. 95). Competitive intelligence serves three purposes: forecast rivals’ strategies, predict their reactions to your moves, and work out how to influence their behaviour. The boundary with industrial espionage is not always clear (pp. 95–96).

The problem is usually too much information, not too little. Michael Porter’s four-part framework says what to look for (Grant, 2018, p. 96). Each part feeds the same two predictions: what will the competitor do next, and how will it respond to us?

Allocentrism in practice

Kiwi had to see the game from the major airlines’ side to pick a capacity they would accommodate; BellSouth turned a weak hand into a payment because it understood what Lin and McCaw wanted (Brandenburger & Nalebuff, 1995, p. 71). A competitor profile is the same discipline written down.

01

Future goals

Financial or market goals? A rival chasing share fights harder than one chasing profit, and one free of profit discipline, such as a state-owned firm, is hardest of all. The further performance falls short of its goals, the likelier a change of strategy, perhaps of management (p. 97).

02

Current strategy

How is it competing now? Look at what the company says and at what it does, and reconcile the two (p. 96). A forecast of change starts from an accurate picture of the present.

03

Assumptions

What its managers believe about the industry and about themselves. These beliefs are stable and converge across an industry into a shared “recipe” that can blind a whole industry. British and US motorcycle makers dismissed light Japanese bikes in the 1960s; by 1980 BSA and Triumph had stopped production and Harley-Davidson was struggling to survive (p. 97).

04

Resources & capabilities

Strengths and weaknesses decide how serious a challenge or a reply can be. Do not start a price war with a rival sitting on a cash pile; aim at weaknesses it will find hard to answer (p. 98). Session 5 builds the tools for this part.

Activity · competitor profile

Profile GM as Tesla’s strategists might have in 2018

Eight pieces of evidence from the cases. File each under one of Porter’s four components and check. Some fit two; the feedback says which. Then use the whole profile to predict GM’s response to the Model 3.

0 of 8 checked

Your prediction

Given the profile, what was GM’s most likely response to Tesla’s mass-market push?

Question 03 · Courtney, Kirkland & Viguerie (1997)

What should a strategist do when the future cannot be forecast?

The article opens on two firms betting on digital photography in opposite ways: Kodak spending $500 million a year on digital products, Hewlett-Packard $50 million a year on a rival vision built around home photo printers. Most companies, the authors note, lack the position, assets or appetite for risk to make such industry-shaping bets work; others hedge with small investments that may reserve the right to play or only the right to lose; others wait and see, and open a window for competitors (Courtney et al., 1997, p. 2).

The level of residual uncertainty decides which tools work. The standard kit of market research, competitors’ cost and capacity, the value chain, Porter’s five forces and a discounted-cash-flow model suits level 1 (p. 6). It misleads above it. One of history’s great underestimates makes the point: in 1977 Digital Equipment’s president said there was no reason for anyone to have a computer at home, when the personal-computer boom was already within the range experts were discussing (p. 3).

“Shaping strategies can fail, so the best companies supplement their shaping bets with options that allow them to change course quickly if necessary.”
Courtney, Kirkland & Viguerie (1997), p. 11

The article summarises its argument in exhibits on the four levels, the three postures and the portfolio of actions. This page restates their ideas in cards; read the exhibits themselves in the course pack.

LEVEL 1

A clear-enough future

One forecast is precise enough to point to a single direction; the residual uncertainty does not change the decision. An airline answering a low-cost entrant at its hub can learn segment sizes, the entrant’s costs and capacity, and its objectives (pp. 3–4).

Tools: the standard kit and a DCF (p. 6).

LEVEL 2

Alternate futures

A few discrete outcomes; analysis can suggest probabilities but not pick the winner, and the best strategy depends on which occurs. Pending legislation is one source; another is a rival’s capacity decision in an oligopoly such as pulp and paper (p. 4).

Tools: a valuation model per scenario, decision analysis, game theory, and the trigger variables that signal which future is coming (pp. 5–7).

LEVEL 3

A range of futures

The outcome can fall anywhere on a continuum bounded by a few key variables, with no natural scenarios: a consumer-goods firm entering India might face customer penetration anywhere between 10% and 30%, or the cost of an unproven technology (p. 4).

Tools: no more than four or five scenarios, each distinct, spanning the probable range; test robustness (p. 7).

LEVEL 4

True ambiguity

Several uncertainties interact so that not even the range can be identified. Rare, and transitional: Russia in 1992, consumer multimedia in the 1990s. Such situations drift towards level 3 or 2 (pp. 4–5).

Tools: list what is known and knowable, analogies, indicators to track, and what you would have to believe (pp. 7–8).

Three postures · pp. 8–9

  • Shape. Drive the industry towards a new structure of your own devising, by shaking up a stable industry or by steering a volatile one.
  • Adapt. Take the structure and its evolution as given; win by choosing a position, or by recognising and answering developments fast.
  • Reserve the right to play. A special form of adapting, only at levels 2–4: incremental investments now that give a privileged position in information, cost or relationships, so you can wait until uncertainty falls.

Three kinds of move · pp. 9–10

  • Big bets. Large commitments, such as major capital investments or acquisitions, that pay off big in some scenarios and lose big in others. Shapers usually need them.
  • Options. Modest first investments that secure the upside and limit the downside: pilots, limited joint ventures, licences on an alternative technology.
  • No-regrets moves. Moves that pay off whatever happens: cost cuts, competitive intelligence, skills, and sometimes even capacity or market entry.

A shaper raises uncertainty

FedEx entered a stable, level 1 delivery industry and created level 3 uncertainty for itself and level 2 for UPS: would overnight delivery work, and would UPS have to match it? Its big bets took it to the brink of bankruptcy in its first two years, then reshaped the industry (pp. 10–11).

Shapers can still adapt

Microsoft’s proprietary MSN network was a big bet on one of two futures. When trigger variables showed that open networks would win, it refocused MSN on the internet. Postures are not carved in stone (p. 11).

Credibility can replace cash

At levels 3 and 4 a shaper needs the credibility to coordinate other players around its preferred outcome more than deep pockets. Netscape shaped browser standards on its leaders’ credibility (p. 13).

The right to play, or to lose?

At level 4 it is hard to tell whether small investments reserve the right to play or only the right to lose. Prefer cheap options, and re-evaluate them whenever an important uncertainty clears, at least every six months (pp. 13–14).

The trap this session sets

Tesla’s history is easy to read as a string of brilliant moves, because we know how it turned out. Courtney’s levels and Brandenburger and Nalebuff’s demand to see the game from the other players’ side both force the harder question: what did each move look like before anyone knew how the game would end, and what did the other players believe at the time?

Activity · uncertainty classifier

What level of residual uncertainty is this?

Seven decisions, each seen from the date it was made. Ask what would remain unknown after the best possible analysis, and whether the possible outcomes are one, a few, a range, or not even a range. One is not residual uncertainty at all. The boundary between levels 3 and 4 is a judgement; the feedback says when two answers are defensible.

0 of 7 checked

Video · miscellaneous videos on Steve Jobs

Watch: Steve Jobs on rivals, bets and advantage

The course outline pairs this session with videos on Steve Jobs. The three below are the Jobs videos listed in the course textbook’s instructor resources, Rothaermel’s supplementary lecture slides. Each loads YouTube’s privacy-enhanced player only when you press play. The questions are prompts for watching with today’s three questions, not summaries of what is said.

Question 1 · substitutor and complementor

Steve Jobs and Bill Gates: the 2007 interview

Two founders on one stage. Draw Apple’s value net with Microsoft in it: note each point where one firm’s products make the other’s more valuable, and each point where they compete. Is Microsoft a substitutor, a complementor, or both, like American and United buying planes?

Listed in Rothaermel’s chapter 4 slides. Third-party upload. Open on YouTube

Question 2 · competitor analysis, from the other side

Steve Jobs in 1980: rare presentation footage

Sit in an incumbent’s chair. From what Jobs presents, write Apple’s goals, current strategy, assumptions and capabilities in one line each. What would you have predicted Apple would do next, and which of your own assumptions might have blinded you?

Listed in Rothaermel’s chapter 4 slides. Third-party upload. Open on YouTube

Question 3 · deciding without knowing

Steve Jobs’ 2005 Stanford Commencement Address

Jobs tells his career as a series of decisions taken before he knew how they would turn out. For each one, decide what level of uncertainty he faced, and whether he was making a big bet, buying an option or waiting.

Stanford’s own channel; listed in Rothaermel’s chapter 7 slides. Open on YouTube

Five minutes in pairs · claim and credibility

Pick one video. Separate a memorable line from evidence of a strategic commitment. Then ask what would make the promise credible to a supplier or a rival: an investment that is hard to undo, a contract, or a reputation already earned? Brandenburger and Nalebuff would call the line a tactic; Grant would ask whether it removes any options.

Third-party uploads can be taken down. If one will not play here, try its YouTube link, and mention it in class.

The case · Elon Musk at Tesla (2024)

Arguing all three questions on one company

Bring the case having formed a view, not a summary. These prompts are written for this session and map onto the three questions above: the case supplies the evidence, and you supply the argument.

Note on the case. “Elon Musk at Tesla” (Serafeim & Migdal, 2024) is a Harvard Business School case and is not reproduced here; obtain it through the course pack. Sull and Reavis (2019) is a supplementary companion case on Tesla’s entry. The prompts and classifications below are original to this session; full references are at the end.

Against Q1

Draw Tesla’s value net in 2019 and again in 2021. Which players hold two roles? Daimler bought Tesla’s battery packs in 2009 and appears as a rival in Exhibit 11b; Panasonic supplied cells and co-funded the Gigafactory (Serafeim & Migdal, 2024, pp. 6, 9). Then find the player with the most added value in each version, and say whether Tesla’s profit followed it.

Against Q2

Pick the one Tesla move that most changed the game, not just won it. Name the PART it changed, the player it hurt, and that player’s best reply. Was the move win–lose or win–win for the incumbents, and would their imitation have helped or hurt Tesla?

Against Q3

Take Gigafactory 1, the SolarCity purchase and Optimus. For each, say what level of uncertainty Tesla faced when it moved, whether the move was a big bet, an option or a no-regrets move, and what trigger variable should have told Tesla to expand, stop or change course.

Tesla’s moves, classified: posture and portfolio

Courtney’s FedEx story has a Tesla shape. In 2003 the industry looked stable: carmakers spent three to five years and hundreds of millions or billions of dollars on each platform, sold through dealer networks, and ran on an infrastructure of petrol stations (Serafeim & Migdal, 2024, pp. 3–4). Musk’s reading in 2003 was that the industry ran on two false premises: that a compelling electric car could not be built, and that no one would buy one (p. 4).

A shaper in a level 1 industry raises uncertainty, for itself and for the incumbents (Courtney et al., 1997, p. 10). Classify each move below by the posture it served and the kind of move it was, as it looked when it was made. One incumbent move is included for contrast.

The financial backdrop · Serafeim & Migdal (2024)

  • 2003–2010: $290 million of losses on $150 million of revenue (p. 7).
  • End of 2009: under $500,000 of operating cash (p. 6).
  • End of 2016: $1.4 billion of free cash, burning about $500 million a quarter (p. 10).
  • FY2021: EBIT margin 12.1%, above GM’s 9.0% and Toyota’s 8.1% (Exhibit 11b).

Activity · posture and move

Shape, adapt or reserve? Big bet, option or no-regrets?

Eight moves, in date order. Pick one answer in each row, then check. The score counts each row separately.

0 of 8 checked

The pattern

Tesla’s portfolio is a shaper’s. The early moves were options and no-regrets moves that kept it alive and taught it the market; the big bets went into the ecosystem pieces that passed the bottleneck test: charging and battery supply. Note when the biggest bets came. Tesla bought SolarCity in November 2016, the year it ended with $1.4 billion of free cash while burning about $500 million a quarter (Serafeim & Migdal, 2024, p. 10). FedEx’s big bets took it to the brink in its first two years (Courtney et al., 1997, pp. 10–11).

The incumbents’ portfolio

By 2018 the incumbents were buying positions as well as cars: GM a 9% stake in Lyft, Toyota talks with Panasonic on a battery joint venture, BMW a $1.1 billion contract with CATL for a battery factory in Europe (Sull & Reavis, 2019, pp. 7, 10). Which of these reserve the right to play, and which only the right to lose? What trigger variable would tell you, and how often would you look?

The question the room usually splits on

Was Tesla a game maker or a game taker? One reading: Tesla changed the players (no dealers), raised its added value at the bottlenecks (cells, charging, software) and shaped a level 1 industry into a new one. The other: governments changed the rules for Tesla, with tax credits, emissions limits and combustion-car bans, and much of its early profit was a regulatory credit paid by other carmakers. Both readings fit the evidence, and they imply opposite things about durability. Added value survives a change in the rules; a rule can be rewritten by others.

Take a side before class. The exhibit workshop below gives you the numbers; you will be asked which figure would change your mind.

Discussion questions

Come with a written answer to the first two. The others are for the room.

  1. 01

    Tesla built its Supercharger network itself and kept it for Tesla drivers; BMW and Nissan backed EVgo’s fast chargers, open to every EV driver (Sull & Reavis, 2019, pp. 7, 14). Use the bottleneck test and added value: which choice lets its owner capture more value, and which grows the EV market faster? On what terms should Tesla open its network, and to whom?

  2. 02

    Build the competitor profile for an incumbent other than GM, using only the cases. Predict its response to the Model 3, and name the one piece of evidence that would most change your prediction.

  3. 03

    Regulatory credits were 2.4% of Tesla’s revenue in 2019 but 14.6% of its gross profit (Serafeim & Migdal, 2024, Exhibits 11a, 11c). Who is paying, where does that sit in Tesla’s value net, and what happens to Tesla’s added value in that market as the incumbents’ own EV sales grow?

  4. 04

    Grant says game theory explains the past better than it predicts the future. Explain the 2014 patent offer with PARTS. Then say what, if anything, looking forward and reasoning backward could have told Tesla about it in advance.

  5. 05

    One reading in the case calls Musk’s 2018 pay plan a signal of credible commitment. In Grant’s sense a commitment removes options; in Brandenburger and Nalebuff’s it is a tactic that changes perceptions. Which options did the pay plan remove, and which did the Gigafactory remove?

  6. 06

    In 2024 Musk said he wanted about 25% voting control before growing Tesla into a leader in AI and robotics (Serafeim & Migdal, 2024, p. 15). What level of uncertainty does that business face, and what posture and portfolio of moves would Courtney et al. recommend for it?

The exhibit workshop · 20 minutes

Is Tesla’s profit earned, or paid for by its rivals?

It is early 2022. You sit on the investment committee of a fund that holds Tesla shares, priced at 216 times earnings against 8.7 for GM (Serafeim & Migdal, 2024, Exhibit 11b). A colleague says: “Tesla’s profit is regulatory credits that other carmakers pay for. Strip them out and it is a carmaker like any other.” Tesla earns the credits by beating emissions rules and sells them to other regulated firms (Tesla, 2022). Decide whether your colleague is right.

Figures are from the case’s Exhibits 11a, 11b and 11c, in percent of revenue unless marked. Compute before you reveal: the reveal is worth nothing if you have not committed to an answer first. Calculations assume, as the exhibits imply, that credits carry almost no cost.

Exhibit A · Tesla, FY2019–FY2021 (Exhibits 11a, 11c)

YearRevenue ($bn)Gross margin EBIT marginCredits, % of revenueCredits, % of gross profit
FY201924.616.6%0.3%2.4%14.6%
FY202031.521.0%6.2%5.0%23.8%
FY202153.825.3%12.1%2.7%10.8%

Task 1. Credits’ gross profit in points of revenue = share of gross profit × gross margin. Compare it with credits’ share of revenue. What is the credits’ gross margin?

Task 2. Subtract those points from the EBIT margin. What would Tesla have earned at the operating line without credits in each year?

Exhibit B · Tesla and selected rivals, FY2021 (Exhibit 11b)

TeslaGMVWFordHondaToyotaMercedes
Gross margin25.3%14.2%17.8%12.1%20.7%17.8%21.8%
EBIT margin12.1%9.0%8.0%7.8%5.0%8.1%11.2%
Price/earnings2168.78.74.78.710.73.1

Task 3. Put Tesla’s FY2021 EBIT margin without credits from task 2 into this table. Where does it rank among the seven?

Exhibit C · Two complements in Tesla’s own accounts, FY2017–FY2021 (Exhibits 11a, 11c)

FY2017FY2018FY2019FY2020FY2021
Total gross margin18.9%18.8%16.6%21.0%25.3%
Services & other, % of revenue8.5%6.5%9.1%7.3%7.1%
Services & other, % of gross profit−10.3%−12.1%−13.4%−5.5%−0.8%
Energy generation & storage, % of revenue9.5%7.2%6.2%6.3%5.2%
Energy generation & storage, % of gross profit10.9%4.7%4.7%0.3%−0.9%

Task 4. Each line’s own gross margin = its share of gross profit × total gross margin ÷ its share of revenue. Which of these complements makes money?

Only after you have written your own.

What the numbers do and do not settle

Read the credits through the value net. Tesla sells them to other firms regulated under the same vehicle-emissions rules (Tesla, 2022), which makes the carmakers that buy them Tesla’s substitutors in the car market and its customers in the credit market. Brandenburger and Nalebuff would ask what Tesla’s added value is in that second game. It is high while rivals sell few EVs of their own, and it falls as their own EV sales let them meet the rules themselves. A rule, they warn, can be rewritten by others.

What the exhibits cannot settle: whether credits really carry no operating cost (the EBIT figures above assume so), whether the losses on service were a deliberate subsidy to sell cars or simple inefficiency, and how long the credit market lasts. Those are Courtney questions: name the level of uncertainty before you value the stock.

Name the payer

Which player in Tesla’s value net funds each line of profit?

Name the rule

Which part of the profit depends on a rule Tesla does not write?

Give the number

What figure would change your vote on the investment committee?

In the room

Workshops & simulation

Two decisions with the outcome revealed, and a branching simulation that runs all three questions in sequence. The workshops use case facts where they are cited; every outcome and every step of the simulation is a stylised scenario written for this session, not a report of what happened.

In the room

Studio: two rounds, one decision each

Round 1 · 12 min · pairs

One ecosystem move, written as an option

Use the 2024 Tesla case to pick one concrete ecosystem move: the Supercharger network, the Panasonic partnership, the Hertz order, the regulatory credits or Optimus. Classify its primary PARTS lever. Name one complementor that could also compete with Tesla, and use the case to say what that player wants.

Then write the next step as an option, not a wish, in four parts: pilot → signal → threshold → action. What small commitment buys the right to expand? What observable signal would you watch? At what level would you expand, stop or change? And what does keeping the option open cost?

What to hand in

One line: move / actor / incentive / case evidence. A second line: the threshold that would trigger action, and the cost of waiting for it.

Round 2 · 15 min · fours

Role play: the charging-access negotiation

Tesla proposes to let another carmaker’s drivers use its Superchargers if the carmaker shares the cost (Sull & Reavis, 2019, p. 14). Roles: Tesla’s head of charging (sets the terms); the rival’s EV chief (wants access without dependence, and has a shared network such as EVgo as an alternative); a site owner (hosts the chargers and wants a share); a red-teamer who plays the strongest affected player and must name its best counter-move. Every proposal must name the PART it changes.

After the red-teamer speaks, Tesla revises its proposal once. Look forward and reason backward: what does each player do after the deal, not just at it?

What to hand in

The revised terms, the counter-move that forced the revision, and whether the deal is win–win or win–lose for the rival.

Last 12 minutes — the in-class quiz (graded)

The last 12 minutes of every class are a graded quiz, taken individually. Today’s covers the three questions and the case: the value net and added value, PARTS and the game-theory vocabulary behind them, Courtney’s four levels with their postures and moves, and their application to Tesla. To prepare, make sure you can draw Tesla’s value net from memory, give one example of each PART from the article, and place a decision at a level of uncertainty with a reason. The final quiz and concept check below are practice; the graded questions are not on this page.

Knowledge check

Final quiz: fifteen questions

Score: 0 / 0

Test your grip on the session’s core ideas, including the workshops and simulations above. Each answer explains itself.

Concept check

Twelve questions, with reasons

0 / 0 correct

Answer first; the reasoning appears immediately. A wrong answer here is cheaper than a wrong answer in the exam.

See it · explain it · test it

Change the game, then test the response

Cooperation can enlarge total value while competition still determines who captures it. A strategic move needs a credible reaction from other players and a way to learn under uncertainty.

Change the game, then test the responseBegin with the present value net. Propose a specific change to Players, Added value, Rules, Tactics or Scope. Predict other players’ responses rather than assuming they remain still. Under uncertainty, choose a learning move, observe an explicit trigger and decide whether to expand, revise or stop. The diagram is a reasoning sequence, not a forecast.Value net todayWho contributes and who can leave?One PARTS leverChange players, rules or another leverLikely reactionWhat would each player do next?Learning decisionPilot, trigger and next commitment
Original teaching diagram. Read the same reasoning in the numbered explanation beside it; arrows are propositions to test.

Read the diagram as a story

Begin with the present value net. Propose a specific change to Players, Added value, Rules, Tactics or Scope. Predict other players’ responses rather than assuming they remain still. Under uncertainty, choose a learning move, observe an explicit trigger and decide whether to expand, revise or stop. The diagram is a reasoning sequence, not a forecast.

  1. Current game: fictional BridgeCharge has chargers; local cafés have locations and customers; EV drivers need charging and a place to wait.
  2. Lever: invite cafés into the value net as location partners. This changes Players; a negotiated service agreement may also change Rules.
  3. Reaction: cafés may ask for a larger share, rivals may bid for the same sites and drivers may prefer another route. More partners do not guarantee more retained profit.
  4. Learning: begin with a reversible pilot, track repeat use and contribution after partner payments, and specify the conditions for expansion.

One concept, three business contexts

Hypothetical · original teaching scenario

BridgeCharge: make the café a complement

A fictional charging operator approaches roadside cafés. Charging could bring visitors; a comfortable stop could make charging more attractive. That is a complement hypothesis, not proof of demand. The operator proposes a small pilot and a contract allowing a later expansion decision. The café’s share and the rival’s counteroffer matter: creating a larger pie does not tell you who will retain the extra value.

Discuss: Which PARTS element changes first? What concession would make the move unattractive?

Global · historical example: GM’s 1992 card

General Motors: imitation need not destroy every benefit

Brandenburger and Nalebuff describe GM’s reward card, launched with Household Bank in 1992, and subsequent rival card programmes. Their argument illustrates how a move can alter incentive competition and affect other players’ responses. This is the article’s historical interpretation, not a claim about today’s automobile market. Use the example to ask whether copying a move merely redistributes customers or changes the rules and incentives of the game.

Discuss: Under what assumptions could a copied move improve the game? Under what assumptions would it intensify a price war?

Brandenburger and Nalebuff (1995), historical discussion in the course pack

Indian · NPCI’s published UPI product model

UPI: common infrastructure and competing applications

NPCI’s product booklet describes UPI access to multiple bank accounts through one application and merchant payment functionality. Teaching inference: interoperable infrastructure can support relationships among banks, applications, merchants and users while applications still compete for usage. Keep the focal perspective explicit. Merchant participation may complement an app’s value to users; two apps may simultaneously share infrastructure and compete. Payment volume alone does not establish one app’s profitability.

Discuss: For a payment app, which actors are complementors and which are substitutors? Could either role change?

NPCI UPI product booklet, benefits and ecosystem description

Four diagnostic decisions

Choose an answer, read the reason and compare it with the diagram. Start a fresh attempt when you want to revisit all four decisions.

Practice score: 0 / 4.

    Discuss, then defend a decision

    Argue from the other player’s chair

    Write BridgeCharge’s offer, then let a second pair act as the café. State that player’s alternatives and counteroffer. Revise the original recommendation only after considering the response.

    A genuine option, or a small irreversible bet?

    Specify the right a pilot preserves, its cost, a learning indicator and the decision it informs. Explain why a positive pilot result may fail to transfer to the next set of sites.

    Use the supplied resources with this lab

    The guide below maps the supplied filenames to their teaching purpose. Read the authorised course-pack originals alongside the existing reading list; the lab provides original explanations and practice.

    • Brandenburger-Nalebuff-HBR-1995.pdf

      The Right Game (1995): use the value net and PARTS discussion to state what changes, and the added-value reasoning to distinguish contribution from total value. Read the GM example as a historical argument, then test its assumptions. Obtain the original through the course pack.

    • Strategy-Under-Certain.pdf

      Despite this filename, the supplied article is Strategy Under Uncertainty by Courtney, Kirkland and Viguerie (1997). Diagnose uncertainty remaining after analysis, then choose posture and moves. A pilot is an option only when it preserves a useful later choice. Course-pack access remains the route to the original.

    • Tesla case and existing video prompts

      Continue the assigned Elon Musk at Tesla case and the current exhibit workshop. Use the existing Watch section’s Steve Jobs questions as evidence prompts, and the textbook for formal game-theory tools. These activities remain in their original locations.

    Return to the learning outcomes · Continue the existing workshops · Open the complete reading list

    Preparation

    Readings for Session 4

    Read Brandenburger and Nalebuff first, then Courtney, Kirkland and Viguerie, and reread the case last: it is the material you argue with, and the articles are the tools. Readings and cases are obtained through the course pack; none is reproduced or linked here.

    1. Case

      Serafeim, G., & Migdal, A. (2024). Elon Musk at Tesla (HBS Case No. 9-123-044). Harvard Business School.

      Carried over from Sessions 1–3. Reread pp. 5–15 for Tesla’s moves and Exhibits 11a–11c for the exhibit workshop.

    2. Required · Q1 and Q2

      Brandenburger, A. M., & Nalebuff, B. J. (1995). The right game: Use game theory to shape strategy. Harvard Business Review, 73(4), 57–71.

      The value net, added value, coopetition and PARTS. Bring one PARTS proposal for Tesla and one credible response from another player.

    3. Required · Q3

      Courtney, H., Kirkland, J., & Viguerie, P. (1997). Strategy under uncertainty. Harvard Business Review, 75(6). (Reprint 97603)

      Four levels of residual uncertainty, three postures, three kinds of move. Page numbers on this page follow the reprint’s own pagination (pp. 1–15) in the course pack.

    4. Supplementary · historical companion case

      Sull, D., & Reavis, C. (2019). Tesla’s entry into the U.S. auto industry (Case No. 18-186). MIT Sloan School of Management.

      Reread pp. 5–14: dealers, charging, connectivity, mobility services, competitors’ EV plans, the Gigafactory and Tesla’s sales and service model. Licensed CC BY-NC-ND 3.0.

    5. Further · the textbook chapter

      Grant, R. M. (2018). Contemporary strategy analysis (10th ed., ch. 4, pp. 83–106). Wiley.

      “Further topics in industry and competitive analysis”: ecosystems, business models, game theory and competitor analysis. The source of the formal game theory on this page; the 11th edition’s Indian adaptation keeps the chapter.

    6. Further · platforms and business models

      Rothaermel, F. T. (2024). Strategic management (6th ed., sections 7.5 and 12.3). McGraw Hill.

      The why/what/who/how questions for a business model. Its platform roles lead to Session 8, and its catalogue of business-model types to Session 13.

    7. Video

      Miscellaneous videos on Steve Jobs

      The three in the Watch section above. Watch at least one before class, with its question in mind.

    8. Carried forward

      Porter, M. E. (2008). The five competitive forces that shape strategy. Harvard Business Review, 86(1), 78–93.

      Session 3’s reading: today starts where its section on complements stops. Porter (1996), Session 1’s reading, supplies the idea of fit used in the business-model table.

    Earlier interactive decks

    Their content is merged into this page. They were built for other course numberings, so the session numbers in their own titles differ from this page’s.

    References

    APA 7th edition. Cited in the text by author and date.

    1. Brandenburger, A. M., & Nalebuff, B. J. (1995). The right game: Use game theory to shape strategy. Harvard Business Review, 73(4), 57–71.
    2. Courtney, H., Kirkland, J., & Viguerie, P. (1997). Strategy under uncertainty. Harvard Business Review, 75(6). (Reprint 97603)
    3. Grant, R. M. (2018). Contemporary strategy analysis (10th ed.). Wiley.
    4. Porter, M. E. (2008). The five competitive forces that shape strategy. Harvard Business Review, 86(1), 78–93.
    5. Porter, M. E. (1996). What is strategy? Harvard Business Review, 74(6), 61–78.
    6. Rothaermel, F. T. (2024). Strategic management (6th ed.). McGraw Hill.
    7. Serafeim, G., & Migdal, A. (2024). Elon Musk at Tesla (HBS Case No. 9-123-044). Harvard Business School.
    8. Sull, D., & Reavis, C. (2019). Tesla’s entry into the U.S. auto industry (Case No. 18-186). MIT Sloan School of Management.
    9. Tesla, Inc. (2022). Form 10-K for the fiscal year ended December 31, 2021. U.S. Securities and Exchange Commission. sec.gov