Deterrence, preemption, and a company handed to the planet
Session 6 asked how a firm opens a gap between what customers will pay and what it costs to serve them. This session asks what keeps that gap open once rivals have seen it. The evidence is unkind: in one large sample of businesses, more than nine-tenths of the profit gap between the best and the worst had closed within ten years.
Ghemawat and Pisano’s answer is that advantage survives only behind barriers, and that the barriers which last are built from commitments that are hard to undo and capabilities that are slow to copy. Patagonia’s 2022 decision to give its voting stock to a trust and the rest to a nonprofit is the test case: the least reversible move the company has made, read here as strategy rather than as philanthropy.
What this session answers
The case does not use strategy vocabulary, so the reading is ours to make. Facts from the case are marked Case; where this page reads the case as strategy, the card is marked Analysis.
Carried forward · Sessions 1, 5 and 6
You already know how to find an advantage and how to test whether it is valuable, rare and hard to copy. Today the question changes from is there an advantage? to what stops it being competed away, and for how long?
Session 6 · Creating advantage
Ghemawat and Pisano open their note with the standard view of advantage: open the widest possible gap between the benefits delivered to customers and the costs the organisation incurs (1997, p. 1). Session 6 built that wedge for Patagonia. This session asks why wedges so rarely stay wide.
Session 5 · Resources and VRIO
In Session 5 the I in VRIO was one test among four, and isolating mechanisms were the reasons a resource might pass it. Here we ask what makes imitation costly, who pays to keep it costly, and what happens to the value even when nobody imitates.
Session 1 · Porter (1996)
Operational effectiveness diffuses quickly, and fit among many activities is harder to match than any one of them. Ghemawat and Pisano accept that argument (p. 12) and push past it: commitments and capabilities mean rivals do not all face the same productivity frontier (p. 13). Revisit Session 1
The trap this session sets
“Hard to copy” is not the same as “lasting”. An advantage nobody can imitate can still be drained by the people who supply it or wasted inside the firm that owns it. For every barrier on this page, ask two questions: who pays to keep it up, and who could take the value it protects?
The problem · Ghemawat & Pisano (1997), pp. 1–11
An advantage is necessary for superior performance, but far from sufficient. Ghemawat and Pisano’s first example is the real mousetrap industry: spring traps, repeating traps and glueboards all succeeded in the market without making their inventors rich, because they attracted imitators as well as mice (p. 1). Before you read on, commit to a number.
Activity · predict, then reveal
The PIMS database held a complete ten-year record, 1971 to 1980, for 692 business units. Split them into two equal groups by 1971 return on investment: the top group averaged 39%, the bottom group 3%. Keep every business in the group it started in. By 1980, how much of the 36-point gap between the two averages had closed (Ghemawat & Pisano, 1997, p. 1)?
| Measure | Value | Source |
|---|---|---|
| Business units with a complete 1971–80 record | 692 | p. 1 |
| 1971 ROI, top half / bottom half | 39% / 3% | p. 1 |
| Share of the spread managers usually guess had closed | one-third to one-half | p. 1 |
| Share that actually closed by 1980 | more than nine-tenths | p. 1 |
The note’s Exhibit 1 plots the yearly values; this page shows only the figures stated in its text.
Fruhan followed large companies whose return on equity averaged more than 25% in 1976–82. Their median ROE ran 21% above the S&P 400 average in that period, but only 2% above it in 1989–93 (Ghemawat & Pisano, 1997, p. 2). A McKinsey study that defined excellence as ROE plus sales growth reached the same verdict about its impermanence.
A business earning 39% would be unwise to demand 39% from every new investment, so part of the fall is value-maximising choice. Ghemawat and Pisano argue that by far the larger part is unwanted and even unanticipated (p. 2). Managers understand regression to the mean; they misjudge how far and how fast it runs (p. 1).
A superior position pays off over time only if it keeps two properties: scarcity (rivals cannot simply have the same thing) and appropriability (the owners, not someone else, pocket the value). Each is threatened twice (pp. 2–3).
Scarcity · threatened from outside
Rivals copy the position.
Competitors learn the details of most new products within a year; imitation tends to cost about a third less than innovation and to be about a third quicker (p. 3). As filters took over 80% of the US cigarette market, varieties rose from 18 in 1951 to more than 80 by 1967, and the advertising-to-sales ratio from 3.8% to 8.7% (p. 4).
Rivals go around the position.
Changes in technology, demand and input prices, in that order, open the way (p. 6). In PBX switches AT&T harvested its old base, Rolm spent to defend its position, and InteCom skimmed the market and recouped ten to a hundred times its investment (p. 7). Sometimes the right move is to take the money and run.
Appropriability · threatened from inside the value chain
Non-owners divert the value.
How complementary a party is sets the ceiling; its best alternative to a deal sets the floor (p. 8). NFL team revenue rose 77% in real terms in 1970–84, yet operating income fell by a third: players’ share of revenue went from 35% to nearly 55% (pp. 8–9).
The value is wasted inside.
Estimates put it at 10% to 40% of revenues, and the richest positions leave the most room for it (p. 10). By the late 1970s Xerox was dissipating at least 20% of its sales: about $5 billion over five years, more than its operating income for the period (p. 10).
All four threats grow with the value a position creates, and all four take time to bite (p. 3). That is why a fat margin is not evidence of a durable one.
Analysis Keep the four for the case. Patagonia’s 2022 ownership change says little about imitation or substitution, and a great deal about who can divert or dissipate the value Patagonia creates.
The outline’s question · Grant (2019), ch. 7, pp. 162–165
Grant borrows Richard Rumelt’s term isolating mechanisms for the barriers that stop superior profitability being competed away, and organises them around the steps an imitator must take. To copy a rival, a firm must identify the advantage, have an incentive to imitate, diagnose what causes it, and acquire the resources and capabilities to reproduce it. At each step the incumbent can put a barrier in the way (Grant, 2019, p. 162).
STEP 1 · IDENTIFY
Barrier: obscure superior performance
A private company need not disclose its results. Few food processors saw how profitable pet food was until the UK Monopolies Commission revealed that Mars’s Pedigree Petfoods earned a 47% return on capital employed (p. 162). Limit pricing, setting price just below the level that attracts entry, gives up short-term profit for the same end (p. 163).
STEP 2 · INCENTIVE
Barrier: deterrence and preemption
Persuade rivals that imitation will not pay, or occupy the niches before they arrive (p. 163). Both work on the rival’s expected payoff rather than its ability to copy, and both are taken apart in the next section.
STEP 3 · DIAGNOSE
Barrier: causal ambiguity
When advantage rests on a complex bundle of resources and capabilities, rivals cannot tell which part matters, so imitation is uncertain (Lippman and Rumelt). Where activities are tightly linked, an imitator that gets the configuration slightly wrong can land in the valley next to the peak (pp. 163–164).
STEP 4 · ACQUIRE
Barrier: immobile, hard-to-replicate resources
Having diagnosed the advantage, the imitator must buy or build what it needs. Limits on transferring and replicating resources are the barrier here (pp. 164–165). Session 5’s costly-to-imitate test lives at this gate.
Ghemawat and Pisano list five early-mover advantages that impede imitation (pp. 3–4). Each one works at one or more of Grant’s gates.
| Early-mover advantage | How it impedes imitation | Grant’s gate |
|---|---|---|
| Private information | Knowledge that is tacit, and that no single person can carry out of the firm | Diagnose |
| Size economies | Scale, learning or scope; a rival that matches your size fears supply will outrun demand | Incentive |
| Contracts and relationships | Early movers lock in better terms with suppliers, buyers and channels | Acquire |
| Threats of retaliation | Credible only with both the ability and the willingness to retaliate | Incentive |
| Response lags | Time to observe and to implement; they defer imitation without stopping it | All four, slowed |
The logic worth remembering
An imitator must clear all four gates. The incumbent therefore needs only one gate that genuinely holds. That is the same arithmetic Porter applies to fit in Session 1: a rival copying a system must match every piece, so the defence is as strong as its strongest link, not its weakest.
Set that against the PIMS evidence. Grant reads the other side of the record: profit differences between competitors often persist for a decade or more, which suggests isolating mechanisms can be very effective (2019, p. 162). Both are true; the question is which barriers you have.
Activity · the imitation chain
Imagine a well-funded, shareholder-owned outdoor-apparel company (hypothetical) that wants Patagonia’s position. For each gate, decide whether Patagonia’s barrier holds, leaks or is open. Your rating unlocks our reading, with the evidence marked as case fact or analysis.
Deterrence and preemption · the incentive gate
Both work on the second of Grant’s conditions. Deterrence changes what the rival expects will happen if it moves; preemption removes the opportunity before it can move (Grant, 2019, p. 163). Neither stops a rival that is able to imitate. Both try to make it unwilling.
Deterrence
Grant’s three tests: “For deterrence to work, threats must be clearly signaled, backed by commitment, and credible” (2019, p. 163).
1 · SIGNALLED
The rival knows what will happen if it moves.
2 · COMMITTED
Something already done makes backing down costly.
3 · CREDIBLE
Carrying it out would not hurt you more than the rival.
A threatened price war usually fails the third test, because it would damage the incumbent more than the entrant; capacity already built is more persuasive (Grant, 2019, p. 92). Ghemawat and Pisano are blunter: “Talk of retaliation is, however, cheap” (1997, p. 4). A threat needs both the ability and the willingness to retaliate, shown either because retaliation pays directly or because the firm has already been tough at a loss.
Preemption
Occupy existing and potential niches so the challenger has fewer places to invest (Grant, 2019, p. 163). Grant names three forms:
The cleanest case in the reading · Du Pont
Du Pont’s titanium dioxide process used cheaper feedstock and gave it a 20% cost advantage. Mastering it took $50–100 million and several years of testing in an efficiently scaled plant. Du Pont also built a large new plant that crowded out rivals’ large-scale expansion, and twenty years later it was still the only firm in the world running the low-cost process (Ghemawat & Pisano, 1997, p. 14).
Moving first is no panacea
In four of the five filtered-cigarette segments, a top brand’s order of entry matched its share rank; the odds of that by chance are under 1 in 250. But in plain filters, the largest segment, first-mover Viceroy was overtaken by Winston and then by third-mover Marlboro, repositioned with a cowboy image and a flip-top box (Ghemawat & Pisano, 1997, pp. 4–5). Their conclusion borrows Alfred P. Sloan: “If you are late, you have to be better” (p. 5).
What makes both work
Commitment. A deterrent is credible only if it is backed by commitment, and commitment means eliminating your own options. Grant’s example is Cortés destroying his ships on arriving in Mexico in 1519: there was no going back (2019, p. 92). Preemption is a commitment too: Ghemawat and Pisano credit Du Pont’s position to its aggressive capacity commitments (1997, p. 14). That is the bridge to Ghemawat and Pisano’s main argument, below.
Activity · classify
Eight real moves from the readings. Does each one work mainly by threatening the rival, by occupying the space first, or by doing both at once? Two of them are genuinely both.
Building the barriers · Ghemawat & Pisano (1997), pp. 12–18
The resource view and the activity view both explain why returns might stop converging. Neither, Ghemawat and Pisano argue, explains how a firm gets ahead in the first place (p. 12). Their dynamic view joins them up: the resources a firm has today limit what it can commit to; its commitments shape the activities it performs; and both feed back into the resources it will have tomorrow (pp. 12–13). That gives two ways to build a lasting lead, and both mean rivals no longer face the same efficiency frontier (p. 13).
Resource endowment
fixed in the short run
Resource commitments
route 1: lumpy
Product-market activities
route 2: capabilities
← feedback: what the firm commits to and does today rebuilds the endowment it starts from tomorrow
Route 1 · commitments
“Irreversibility, not the amount of money involved, is the correct measure of whether a particular decision is commitment-intensive or not.”Ghemawat & Pisano (1997, p. 13)
SUNK COST → LOCK-IN
Specialised money that cannot be recovered. At one point Boeing’s net investment in the 747 exceeded its net worth (p. 13).
OPPORTUNITY COST → LOCK-OUT
Options that are costly to revive. Reynolds closed a refinery for good; Pepsi’s late move abroad shows that not investing commits you too (pp. 13–14).
TIME LAG
Years to change course. Coors’s national roll-out needed the better part of a decade (p. 14).
Against the four threats, commitments preempt imitation (Du Pont); they can fight substitution or fund a switch around it (Intel’s commitment to CISC had so far held off RISC); and they move bargaining power against holdup. The American who imported Cuisinart food processors owned the brand, so he could take manufacturing to East Asia and keep most of the value (pp. 14–15). They do least against slack.
Under uncertainty, build scenarios rather than raising the hurdle rate; do not try to stay totally flexible, which brings lock-out and mediocre results; and prefer commitments with a high learn-to-burn ratio, the rate at which you learn whether the commitment is a plum or a lemon divided by the rate at which you sink costs into it. Pilots, staging, milestones and kill triggers raise it (p. 15).
Route 2 · capabilities
A capability is what an organisation can do within a predictable range of proficiency, built over many projects and many years (p. 16). Honda and Toyota in the 1980s aimed the Accord and the Camry squarely at Detroit’s core, at similar prices: no creative positioning. Their edge was developing a comparable model in about half the time with half the engineering resources, and building it for about $2,000 less per car (pp. 16–17).
The catch: superiority is hard to demonstrate, so managers call everything they care about a core capability; coherence is nibbled away by drop-in-the-bucket choices; and specific capabilities bring specific rigidities (p. 18).
The impossibility theorem
No firm can sustain an advantage without superior specialised resources or knowledge. Commitments and capabilities are therefore necessary for sustained success, and not sufficient (p. 18).
Activity · the irreversibility test
Mark which indicators of irreversibility each move shows, then compare with our reading. The first four are the reading’s own examples; the Patagonia decisions are this page’s analysis of case facts; the last is a hypothetical. Where an indicator is not the point of the example, either answer counts.
In-class simulation · Project Chacabuco
It is late 2021 in the role-play. In the winter of 2019 Charles Conn and Kris Tompkins concluded that Chouinard was serious about succession and set up a working group, named after Douglas Tompkins National Park in Chacabuco, Chile (case pp. 6–7). By 2020 Ryan Gellert, Hilary Dessouky and Greg Curtis were on the team, and it met every two weeks for nearly two years (pp. 1, 7). The family has now called the meeting that decides.
A transfer of the stock could trigger a tax of about 40% of the company’s value, easily over $1 billion. With almost all their wealth in Patagonia shares, the Chouinards would have to sell stock to pay it, bringing in outside owners who could dilute the mission (p. 7). The family also wanted to stay involved: Yvon and Malinda were in every day, and both children worked at the company (p. 7).
The goals, as the team understood them · case p. 7
PHASE 1 · 10 MIN
Setting the stage
Read your role card below and fix your character’s priorities.
PHASE 2 · 25 MIN
Proposals and challenges
Conn summarises the options; the family, Tompkins and the lawyers respond.
PHASE 3 · 20 MIN
The breakthrough
Conn names the impasse; Mosley presents; Gellert tests it against operations.
PHASE 4 · 25 MIN
Debate and decision
Argue in role; the family chooses a path below, then reveals.
DEBRIEF
Out of role
Trade-offs, the least-bad option, and what Gellert’s job becomes.
Ten roles plus the facilitator. Open yours; keep the others closed until Phase 2.
Wants serious money for the climate fight, the mission kept beyond your lifetime, and to stay involved. Opposed to a sale or a listing, and to any tax bill that forces a share sale. Wants a model others can follow. Your fear: people saying Patagonia tried this and it did not work (case p. 12).
Shares the mission. Wants the values preserved indefinitely and Yvon kept involved, and worries about the financial stability of the company and its employees under any new structure.
Wants the mission to continue, but not the burden of controlling the company, or of owning assets that large, passed to future generations (case p. 8). Watches the effect on employees.
Wants independence, authenticity and Earth-paced growth, and as much money as possible for environmental work. Does not want the burden of control or of large assets.
Must keep Patagonia profitable enough to fund the mission. Needs to know how dividends will be balanced against reinvestment and pay. Feels “a ton of additional pressure” (case p. 1).
Wants a design for the next fifty years and beyond that releases real money and lightens the family’s load, without selling the company’s soul for a quick cash injection. Opens Phase 2 by summarising the options.
With Doug Tompkins you put the proceeds of The North Face and Esprit into buying and protecting land (case pp. 6–7). Make the team prove that running Patagonia does more for the planet than selling it.
Cares about legal robustness and permanence. Protect the commitments already in the charter, which since 2011 can change only by unanimous shareholder vote (case p. 5).
Owns the tax problem. Find a route that avoids a forced share sale and does not push the family away from the business. Raise the 5% payout rule on foundations in Phase 2.
Expert on large family businesses and foundations (case p. 10). Say nothing about your proposal until the facilitator calls Phase 3, then unlock it below.
Activity · the ownership options
Select an option to see what the case says about it. Argue its score before you show ours. The family then chooses a path and reveals what Patagonia did.
Select an option above.
Patagonia · the 2022 structure as a commitment
On 14 September 2022 Chouinard told an all-staff meeting in Ventura that he was no longer Patagonia’s owner (case p. 1). The structure works because votes and money go to different places. Follow each one.
Activity · follow the flows
Perpetual purpose trust
Patagonia Purpose Trust
Owns 100% of the voting stock, equal to 2% of all shares. Elects the board, approves material transactions and oversees adherence to the company’s philosophy (p. 11).
Receives 2% of dividends, usable only to protect the purpose; no personal use (p. 12).
Irrevocable and in perpetuity; trustees can enforce the purpose but not change it (pp. 10–11).
votes: 100% of voting stock
board elected by the trust
money: 2% of dividends
For-profit company
Patagonia
Existing board and management stay in their roles and the business runs as before (p. 11).
Leadership recommends the dividend from profit left after reinvesting in the business; the board approves it (p. 11). Patagonia estimated $100 million or more a year (p. 12).
money: 98% of dividends
votes: none
501(c)(4) nonprofit
Holdfast Collective
Owns 100% of the non-voting stock, 98% of all shares, and so most of the financial value. Uses its 98% of dividends on the root causes of climate change, through grants and political donations (pp. 11–12).
No voting control (p. 12). Its board includes the four Chouinards (p. 11).
Drawn from the case text, pp. 10–12; the case’s own diagram (Exhibit 7) is not reproduced.
2% / 98%
Voting stock to the trust, non-voting stock to Holdfast (case p. 11).
$17.5M
One-time gift tax on the voting shares, paid by the family (p. 12).
$100M+
Patagonia’s own estimate of annual dividends; an estimate, not a result (p. 12).
$2–8B
Estimated value of a company that did not share its financial data (p. 2).
“While this doesn’t release a large chunk of money up front, just the free cash flow from the dividends over time, it is a better solution than releasing a lot more through a sale of stock or assets and, as a result, selling your soul.” Charles Conn, board chair (case p. 12)
Sunk: the family gave away the economic value of the company and paid $17.5 million to do it. Lock-out: the stock “would not be transferred again” (p. 12), so the sale, the listing and the heir who wants the cash are gone as options. Compare 2011: charter commitments that only a unanimous shareholder vote could change (p. 5) bound a family that held every share (p. 6) only to itself. The limit: the trust could still direct Holdfast to sell some or all of its stock if that served the purpose (p. 12). Hard to reverse, then, not impossible.
Rivals can copy the materials, the repairs and the giving, and Patagonia wants them to: sharing best practice is in its charter (p. 5). What a shareholder-owned rival cannot copy cheaply is an owner that takes nothing out. For that firm it is not an operating decision; its owners would have to give away their claim on the business. The structure blocks Grant’s fourth gate. It is not deterrence: nothing in it threatens a rival.
Grant: “One of the attractions of private company status is avoiding disclosure of financial performance” (2019, p. 162). Patagonia did not publicly share its financial data (case p. 2); a public listing would have meant publishing accounts. With the stock settled for good in a trust and a nonprofit, a listing is now unlikely, so the opacity is locked in as well. The price: customers, staff and the Holdfast Collective also see only what Patagonia chooses to show.
Holdup and slack live in the gap between ownership and control (Ghemawat & Pisano, 1997, p. 8). The new owners cannot take value for themselves: “The purpose trust is the decision-maker, but it doesn’t get any of the financial benefits” (Dessouky, case p. 12). That closes the classic routes of extraction. But the dividend is whatever is left after management’s own reinvestment recommendation (p. 11), and no outside shareholder now prices the firm. Rich positions that are hard to monitor are where slack grows (Xerox, pp. 10–11). The question for Gellert follows: who presses for discipline when the dividend is whatever management says is left after reinvestment, and 98% of it goes to the Holdfast Collective (p. 11)?
Deterrence, preemption, or neither? · Analysis
Read strictly, neither. Nothing in the structure threatens a rival or occupies a niche a rival wanted. It is a commitment aimed inward and forward: at future owners, and at Patagonia’s own managers. If it preempts anything, it preempts Patagonia’s own future options. Argue the other side in the discussion.
Go deeper
Both were built for the 2-year MBA course. Where their facts differ from this page, this page has been checked against the case and the reading; trust it.
For class
Prepare a position on each, with the page of the case or the reading you would cite. The first five use today’s tools; the last asks what they leave out.
01
An imitator must clear all four of Grant’s gates. At which gate is Patagonia’s defence strongest, and did Patagonia build that barrier on purpose or inherit it?
02
Patagonia writes the sharing of best practice into its charter and co-founded an alliance for others to copy its giving. Can a firm invite imitation and still sustain an advantage? Name exactly what is left that a rival cannot copy.
03
Rank the 1994 cotton switch, the 2011 benefit-corporation charter and the 2022 transfer by sunk cost, lock-out and time lag. Which one changed what Patagonia can do, rather than what it promises to do?
04
This page reads the 2022 structure as neither deterrence nor preemption. Make the strongest case that it is one of them. If it is aimed at anyone, who?
05
With no owner able to take money out, who can now divert or dissipate Patagonia’s value? Describe what slack at Patagonia would look like in 2030, and who would notice it.
06
The Tompkinses sold their companies and put the proceeds straight into conservation; Chouinard kept the company and funds the cause from its dividends. “Doug and I argued about this” (case p. 7). Which path fits your own purpose, and why?
Knowledge check
Answer before you look. Each answer explains itself, and the explanation matters more than the score.
Preparation
The reading and the case are copyrighted Harvard Business School Publishing materials. Get them through the licensed course pack; this page does not host or link them.
Required · the reading
Ghemawat, P., & Pisano, G. P. (1997). Sustaining superior performance: Commitments and capabilities (HBS No. 9-798-008). Harvard Business School Publishing.
Read all twenty pages. Pages 12–18, on commitments and capabilities, carry the session.
Required · the case
Trelstad, B., Hsieh, N., Norris, M., & Pinckney, S. (2023). Patagonia: “Earth is now our only shareholder” (HBS No. 9-323-057, rev. September 21, 2023). Harvard Business School Publishing.
Pages 6–12 and Exhibits 7–10 matter most today. Patagonia is also the case for Sessions 5, 6 and 8.
Further · the outline’s framing
Grant, R. M. (2019). Contemporary strategy analysis (10th ed.). Wiley.
Chapter 7, “How is competitive advantage sustained?”, pp. 162–165; chapter 4 on deterrence, commitment and signalling, pp. 91–94.
Further · course textbook
Rothaermel, F. T. (2024). Strategic management (6th ed.). McGraw Hill.
Chapter 4, “Isolating mechanisms: How to sustain a competitive advantage”, pp. 146–151: the Session 5 treatment, worth rereading before class.
Carried forward
Porter, M. E. (1996). What is strategy? Harvard Business Review, 74(6), 61–78.
The fit argument that Ghemawat and Pisano build on.
Patagonia does not publish its accounts. Its revenue (about $1 billion), net income (about $100 million) and headcount (about 3,600) are media estimates reported in the case (p. 2); the $2–8 billion value is an estimated range; the $100 million-plus dividend was Patagonia’s own estimate at the time of the transfer (p. 12). Nothing here reports Patagonia’s performance after 2022, because the course material holds no source for it.