{ "name": "E.I.", "tagline": "From A.I. to E.I.", "version": "1.0.0", "license": "MIT", "source": "https://cei.inc/colab/ei/", "rows": [ { "name": "Ethical", "why": "What will we not do to get there, and can an outsider check what we claim?" }, { "name": "Empathy", "why": "Whose day does this change, and were they in the room when it was designed?" }, { "name": "Ecosystem", "why": "Does it still work after we leave, and who holds the asset at the end?" } ], "columns": [ { "name": "Impact", "why": "Does it work, and for whom?" }, { "name": "Innovation", "why": "Can it be built?" }, { "name": "Investment", "why": "On whose clock, and who owns it at the end?" } ], "center": { "row": "Empathy", "column": "Innovation" }, "marks": { "green": "Evidenced and dated.", "yellow": "A work order with an owner and a date.", "red": "Absent or contradicted. Stops the work until it clears." }, "questions": [ { "n": 1, "id": "measured-not-asserted", "row": "Ethical", "column": "Impact", "pillar": "Purpose", "question": "Can you prove it worked?", "title": "Measured, not asserted", "summary": "A claim arrives with a measure, a baseline, and a date it was taken.", "thesis": "A claim without a measure is a marketing asset. A claim with a measure, a baseline, a date and a named collector is an operating fact. You can check it, argue with it, revise it. The second kind is what we are trying to make, and our own marks say how far we have got.", "requires": [ { "heading": "What the practice requires", "items": [ "The metric is defined in writing before the program starts, denominator included.", "A baseline is captured before the intervention, by someone not judged on the result.", "Every published figure carries its date and its source document.", "Misses publish on the same schedule, in the same place, as hits.", "If a number cannot be collected honestly, we drop the claim rather than soften it." ] }, { "heading": "Why this cell comes first", "items": [ "It makes every other cell auditable. A practice you cannot measure is a preference.", "It anchors to Purpose. You measure to find out whether the thing you said happened.", "Goodhart's problem is real. Once the measure becomes the target it stops measuring. Read Muller alongside Doerr." ] }, { "heading": "Where it goes wrong", "items": [ "Counting outputs and reporting them as outcomes.", "Choosing the denominator after the numerator is known.", "Measuring once, at the moment of maximum flattery." ] } ], "measures": [ "Metric definition dated before program start. Yes or no.", "Baseline captured before intervention. Yes or no.", "Share of published figures carrying a source and a date. Target 100%.", "Claims withdrawn for lack of a measure, per quarter. Zero is a warning, not a win.", "Days from measurement to publication." ], "cases": [ { "verdict": "good", "title": "Patagonia's repair economy", "summary": "A billion-dollar apparel company built on the premise that the most ethical jacket is the one you already own. Repairs and resale run as operations with their own counts, not as a campaign." }, { "verdict": "good", "title": "Warby Parker, fifteen years in", "summary": "Buy-one-give-one was widely dismissed as marketing. Fifteen years of execution held up. The lesson is that the model was checkable, and someone checked." }, { "verdict": "bad", "title": "Theranos and the unfalsifiable demo", "summary": "A $9 billion diagnostics company failed because its core claim never worked. No external measure ever supported it. The absence of the measurement was the product." } ], "reading": [ { "title": "Measure What Matters", "note": "John Doerr, 2018. Objectives with dated, numeric key results." }, { "title": "The Tyranny of Metrics", "note": "Jerry Z. Muller, 2018. What measurement does to the people measured." } ], "paper": "https://cei.inc/colab/ei/papers/measured-not-asserted.html" }, { "n": 2, "id": "hard-stops", "row": "Ethical", "column": "Innovation", "pillar": "Purpose", "question": "Is there a line you will not cross for money?", "title": "Hard stops", "summary": "A written list of harms disqualifies a venture, and nothing offsets them.", "thesis": "Most ethical failure starts as a series of small trades, each one reasonable, where a harm gets priced, offset, and carried forward. A hard stop refuses to price. The list is written and dated before the opportunity arrives. A list written afterwards is an argument.", "requires": [ { "heading": "What the practice requires", "items": [ "The stop list exists in writing, dated and published.", "It is authored before diligence on any specific venture. Amended on a cadence, never in deal week.", "A trigger stops the work. It does not open a negotiation about mitigation.", "The stop is recorded with its reason, so the next team sees the precedent.", "We report the count of ventures stopped. A screen that has never stopped a venture has not been tested." ] }, { "heading": "Why we refuse offsets", "items": [ "An offset assumes the harm and the benefit are in the same unit and land on the same people. They usually are not.", "Permit one offset and the list becomes a price sheet. Then it becomes a budget line." ] }, { "heading": "Where it goes wrong", "items": [ "Writing the list vague enough to survive any deal.", "Putting stop authority inside the team whose pay depends on the deal closing.", "Treating an internal finding of harm as a communications problem." ] } ], "measures": [ "Stop list published, dated, version-controlled. Yes or no.", "Ventures stopped by the list, per year, with reasons recorded.", "Hours from trigger identified to work stopped. Target: same working day.", "Offset transactions permitted. Target: zero, permanently.", "Share of stop decisions made by someone with no economic interest in the venture." ], "cases": [ { "verdict": "ugly", "title": "Purdue and the marketing of an epidemic", "summary": "Aggressive OxyContin marketing contributed to an epidemic linked to over 500,000 U.S. deaths. The information existed the whole time. There was no stop it could trigger." }, { "verdict": "ugly", "title": "The teen-research file", "summary": "Internal research showed Instagram made some teen girls feel worse about themselves. The company chose not to publish it and did not change the product. The stop existed on paper with no authority behind it." }, { "verdict": "ugly", "title": "Palantir's ICE contract", "summary": "FOIA documents showed the software was used in ICE workplace raids, contradicting the company's public description of the contract. A stop list that can be negotiated stops nothing." } ], "reading": [ { "title": "Weapons of Math Destruction", "note": "Cathy O'Neil, 2016. How optimization at scale encodes harm nobody authored." }, { "title": "Ruined by Design", "note": "Mike Monteiro, 2019. The builder's own responsibility for what the built thing does." } ], "paper": "https://cei.inc/colab/ei/papers/hard-stops.html" }, { "n": 3, "id": "ownership-at-the-end", "row": "Ethical", "column": "Investment", "pillar": "Patient Capital", "question": "At the end, who owns the asset?", "title": "Ownership at the end", "summary": "The operator ends up holding the asset, not the payment schedule.", "thesis": "Two financings can move the same money at the same rate and leave the borrower in opposite positions. One ends with an owner. The other ends with a payer. We structure for the first and measure whether it happened.", "requires": [ { "heading": "What the practice requires", "items": [ "Every financing states at signing what the participant owns at term, and on what date.", "Equity accrues from the first payment, not after a qualifying period most never reach.", "Default terms return accrued equity rather than extinguishing it.", "Transfer of title is a tracked deliverable with an owner and a date.", "If the structure cannot deliver ownership, we call it a rental and price it as one." ] }, { "heading": "The precedent", "items": [ "Mondragón was founded in the Basque town of Mondragón by Father José María Arizmendiarrieta and a group of his students. It began as Talleres Ulgor.", "The federation now employs tens of thousands of people.", "Pay ratios between executive and shop-floor work stay tight, bounded by policy rather than sentiment.", "The ownership and the ratio are in the governing documents. That is why they have survived decades of commercial pressure." ] }, { "heading": "Where it goes wrong", "items": [ "Rent-to-own terms designed around a completion rate the seller can absorb.", "Equity that vests only on an event the participant does not control.", "Leverage against an asset the operator has no path to owning." ] } ], "measures": [ "Share of participants holding title or equity at term. Target set per program, published.", "Days from first payment to first dollar of accrued equity.", "Ratio of total payments made to equity accrued, over the life of the facility.", "Residual value to the household on early exit or default.", "Pay ratio inside CEI and inside each venture, published annually." ], "cases": [ { "verdict": "good", "title": "Mondragón", "summary": "A Basque cooperative federation. Worker ownership and bounded pay ratios at the scale of tens of thousands, because both sit in the governing documents." }, { "verdict": "good", "title": "Guild and the tuition that got spent", "summary": "Employer-paid tuition turned from a dusty HR benefit into a real pipeline, by fixing the parts that stop a shift worker enrolling. The worker keeps the credential when she leaves." }, { "verdict": "ugly", "title": "Archegos", "summary": "A family office with no regulatory filings triggered roughly $10 billion of prime-brokerage losses at Credit Suisse and Nomura in a single week. Economic risk without the ownership position that would have made it visible." } ], "reading": [ { "title": "The Citizen's Share", "note": "Blasi, Freeman & Kruse, 2013. The American record on broad-based employee ownership." }, { "title": "Owning Our Future", "note": "Marjorie Kelly, 2012. Ownership design as the variable that decides who benefits." } ], "paper": "https://cei.inc/colab/ei/papers/ownership-at-the-end.html" }, { "n": 4, "id": "first-beneficiary-has-a-name", "row": "Empathy", "column": "Impact", "pillar": "Purpose", "question": "Can you name one real person this helps?", "title": "The first beneficiary has a name", "summary": "One person's day, written down, before anything gets built.", "thesis": "A market is an abstraction. Abstractions do not have Tuesdays. We start each program by writing one real person's day. Their hours, their costs, their workarounds, their words. Every later decision gets checked against that document.", "requires": [ { "heading": "What the practice requires", "items": [ "A named-day document. One interviewed person, their working day, what the problem costs them in time and money.", "Written from first-person interviews with the people it describes.", "Every member of the build team has met a beneficiary in person before capital commits.", "At each gate the solution is read back against the named day. If it no longer fits, the team says which one is wrong.", "The document is versioned, so drift shows." ] }, { "heading": "Interview discipline", "items": [ "Ask about their life, not about your idea.", "Ask what they did last time, not what they would do.", "Take commitment and advance as the only reliable signals. Time, reputation, or money.", "A compliment is not data." ] }, { "heading": "Where it goes wrong", "items": [ "A persona assembled from a deck rather than a person reached by phone.", "Beneficiaries described in aggregate, so no individual can falsify the claim.", "Building for the buyer the team wishes existed." ] } ], "measures": [ "First-person interviews completed before build. A floor, not a target.", "Share of the build team who has met a beneficiary in person. Target 100%.", "Named-day document exists, dated and versioned. Yes or no.", "Gate decisions checked against the named day.", "Advance signals collected. Time committed, money committed, or referrals made." ], "cases": [ { "verdict": "good", "title": "d.light and the end of the kerosene lamp", "summary": "Two Stanford design students designed against a specific household's evening. The cost of kerosene, the hours of light, the smoke. The company's own count runs past 180 million people, and it cash-flows." }, { "verdict": "bad", "title": "Juicero and the $700 bag-squeezer", "summary": "$120 million raised for a Wi-Fi-connected juice press. Bloomberg found the packs could be squeezed by hand. The named day was never written." }, { "verdict": "bad", "title": "Quibi and the 90-second bet", "summary": "$1.75 billion raised, shut down within the year. Nobody had asked whether phones were how people actually watch shows." } ], "reading": [ { "title": "The Mom Test", "note": "Rob Fitzpatrick, 2013. How to ask about the problem without contaminating the answer." }, { "title": "Change by Design", "note": "Tim Brown, 2009. Human-centered design as a working method." } ], "paper": "https://cei.inc/colab/ei/papers/first-beneficiary-has-a-name.html" }, { "n": 5, "id": "built-with-in-the-room", "row": "Empathy", "column": "Innovation", "pillar": "Playbook", "question": "Were the people it affects in the room?", "title": "Built with, in the room", "summary": "Design with them, not for them.", "thesis": "Ecosystem problems have participants who never meet. The operator, the financier, the regulator, the clinician, the engineer. The design defect usually sits exactly where two of them would have had to talk. Putting them in one room for days is the method, which is why the model starts by buying the room.", "requires": [ { "heading": "What the practice requires", "items": [ "Every role the solution depends on is present at each gate, or the gate does not run.", "Sessions are multi-day and residential. A ninety-minute call gets agreement. Design takes days in the same room.", "Decisions are recorded with the roles present. Decisions taken without an affected party are flagged provisional.", "The facilitator is accountable for whether the quiet roles spoke." ] }, { "heading": "Why the campus is a method", "items": [ "The building is the instrument. Multi-day co-presence with overnight accommodation is a capability.", "That is why property sits inside the model rather than beside it. See Own the ground.", "Bring the work to the team, not the team to the work." ] }, { "heading": "Where it goes wrong", "items": [ "Consultation as a review step, after the design is expensive to change.", "The loudest role setting the agenda for the roles with least institutional power.", "Treating co-presence as a nice-to-have, then wondering where the integration defects came from." ] } ], "measures": [ "Stakeholder-days on site per program.", "Distinct roles present at each gate, against the roles the solution depends on.", "Share of gate decisions made with every affected role present.", "Speaking distribution across roles, recorded by the facilitator.", "Campus occupancy by program versus third-party booking." ], "cases": [ { "verdict": "good", "title": "Grameen's second wind", "summary": "Lending terms shaped in villages by the people repaying them. Serious criticism in the 2010s exposed real weaknesses, and the design loop stayed open. It adapted and kept working." }, { "verdict": "bad", "title": "Away and the Glassdoor reckoning", "summary": "A $1.4 billion luggage brand whose culture was designed in a chat tool. The Verge reported it, and leadership learned it from the outside." }, { "verdict": "ugly", "title": "Boeing's 737 MAX", "summary": "A flight-control change that did not require new pilot training. Regulators were not told the system could override the pilot. The people who would fly it were not in the room." } ], "reading": [ { "title": "Sprint", "note": "Knapp, Zeratsky & Kowitz, 2016. Five days to a decision with the right people present." }, { "title": "Team of Teams", "note": "General Stanley McChrystal, 2015. Shared consciousness across organizations not built to talk." } ], "paper": "https://cei.inc/colab/ei/papers/built-with-in-the-room.html" }, { "n": 6, "id": "the-problem-sets-the-clock", "row": "Empathy", "column": "Investment", "pillar": "Patient Capital", "question": "How long will this take, and does the money agree?", "title": "The problem sets the clock", "summary": "The horizon follows the problem, which makes patient money a term structure.", "thesis": "Patience that lives in a founder's disposition evaporates under a capital call. Patience that lives in the term sheet survives one. We state the horizon before capital lands and build the instrument to match.", "requires": [ { "heading": "What the practice requires", "items": [ "Each program states its horizon in writing before the first dollar, with the reasoning.", "The instrument's term matches that horizon. Any mismatch is disclosed to everyone in the deal.", "A program that runs long is re-scoped or closed on evidence. Never force-exited to hit a date that belongs to the capital.", "We report horizon variance. Stated versus actual, across the portfolio." ] }, { "heading": "What patience is not", "items": [ "Not an excuse for the absence of a measure. A long horizon with no interim evidence is drift.", "Not indefinite. A stated horizon arriving with the thesis disproved ends the program.", "Pivot or persevere. Both are decisions. Neither is waiting." ] }, { "heading": "Where it goes wrong", "items": [ "Extrapolating a demand curve from a window that cannot repeat.", "A valuation set by fundraising cadence rather than by the business.", "Exiting when the fund needs a mark, not when the asset is finished." ] } ], "measures": [ "Stated horizon per program, recorded before capital lands.", "Variance between stated and realized horizon, across the portfolio.", "Programs re-scoped or closed on evidence versus force-exited on a date.", "Instrument term versus program horizon. Count of disclosed mismatches.", "Interim evidence checkpoints met on schedule during long-horizon work." ], "cases": [ { "verdict": "good", "title": "The 40-year warranty", "summary": "A warranty and repair operation costed over decades instead of quarters. That turned durability from a marketing position into a unit-economics decision." }, { "verdict": "bad", "title": "Peloton's inventory tide", "summary": "Pandemic demand extrapolated into a capital expansion. It cost the founder the CEO seat and investors billions. The clock belonged to the growth narrative, not the household." }, { "verdict": "bad", "title": "WeWork and the cost of a valuation", "summary": "It peaked at a valuation bigger than most airlines and went bankrupt a few years later, having burned through more investor capital than most funds ever raise. A number produced by a financing cadence and carried as though the business had produced it." } ], "reading": [ { "title": "The Infinite Game", "note": "Simon Sinek, 2019. Playing to stay in the game rather than win the round." }, { "title": "Loonshots", "note": "Safi Bahcall, 2019. Structures that let long-horizon work survive short-horizon pressure." } ], "paper": "https://cei.inc/colab/ei/papers/the-problem-sets-the-clock.html" }, { "n": 7, "id": "measured-at-the-wide-boundary", "row": "Ecosystem", "column": "Impact", "pillar": "Purpose", "question": "Who pays the costs that never reach your P&L?", "title": "Measured at the wide boundary", "summary": "Count where the labor, land, capital and care sit.", "thesis": "Every measurement implies a boundary. The boundary usually gets drawn where the numbers flatter. We draw it first, publish it, then measure. That includes the effects landing outside the entity and inside the ecosystem.", "requires": [ { "heading": "What the practice requires", "items": [ "The boundary is drawn and published before data collection. It includes the labor, land, capital and care the program depends on.", "An externality register is maintained. Effects the program causes that its own P&L does not record.", "Second-order effects are tracked for at least one full cycle after handoff.", "Measurement happens in the field. A result available only on the test rig is a specification." ] }, { "heading": "Systems literacy is the prerequisite", "items": [ "Meadows gives the vocabulary. Stocks, flows, delays, feedback loops.", "Most ecosystem failure is a delay nobody modeled.", "Raworth gives the boundary. Decide what counts as inside before deciding whether you succeeded." ] }, { "heading": "Where it goes wrong", "items": [ "Optimizing the metric inside the app while the cost lands on a driver, a street, a household.", "Measuring under conditions engineered to produce the result.", "Declaring success at handoff. The interesting failures start at handoff plus ninety days." ] } ], "measures": [ "Boundary statement published before measurement. Yes or no.", "Externality register maintained and reviewed on a stated cadence.", "Second-order effects tracked for N cycles after handoff. N stated per program.", "Share of measurements taken in the field rather than on a rig.", "Deployment performance at handoff plus 90 and plus 365 days." ], "cases": [ { "verdict": "good", "title": "How Ecosia planted a forest from ad clicks", "summary": "A small search engine gives away all of its profit and reports more than 200 million trees planted. It accepted a boundary wider than its own margin, then published the counts against it." }, { "verdict": "ugly", "title": "Uber's god-view", "summary": "Internal tools tracked celebrity riders' trips in real time. The company did not see a problem until a journalist did. Metrics improved inside the app while the cost landed outside it." }, { "verdict": "ugly", "title": "Volkswagen's clean diesel", "summary": "Defeat-device software in 11 million vehicles, lowering emissions only during regulatory testing. The boundary was the laboratory. The harm was on the road." } ], "reading": [ { "title": "Thinking in Systems", "note": "Donella H. Meadows, 2008. Stocks, flows, delays, and the leverage points that move a system." }, { "title": "Doughnut Economics", "note": "Kate Raworth, 2017. Drawing the boundary before measuring performance inside it." } ], "paper": "https://cei.inc/colab/ei/papers/measured-at-the-wide-boundary.html" }, { "n": 8, "id": "minimum-viable-ecosystem", "row": "Ecosystem", "column": "Innovation", "pillar": "Playbook", "question": "Does it still work when you walk away?", "title": "The Minimum Viable Ecosystem", "summary": "The smallest set of parts that keeps working after we leave.", "thesis": "An MVP answers whether one thing works. An MVE answers whether the pieces hold together without us. Financed electric vehicles are a product. The vehicle plus financing plus training plus maintenance plus a resale market is the ecosystem. The ecosystem is what survives month nine, or does not.", "requires": [ { "heading": "What the practice requires", "items": [ "Name every role the solution depends on, including the ones nobody is paying for yet.", "Score Adner's two risks before building. Co-innovation risk: what else must succeed. Adoption-chain risk: who else must say yes before the beneficiary benefits.", "Scope the deployment as the smallest version that instruments all named roles at once.", "Then leave, and measure. Survival ninety days after withdrawal is the real result.", "Publish findings either way." ] }, { "heading": "Why a product is not enough", "items": [ "Moore's whole product and Adner's wide lens land on the same point. The customer buys the complete solution.", "The complete solution is assembled from other people's willingness to act.", "The parts that kill ecosystem ventures are rarely the parts the founding team was excited about." ] }, { "heading": "Where it goes wrong", "items": [ "Building the elegant node and assuming the network exists.", "Automating one step of a chain whose other steps still need what the automation removed.", "Calling an integration successful while the team is still in the room holding it together." ] } ], "measures": [ "Roles named and instrumented, against roles the solution depends on.", "Co-innovation and adoption-chain risks scored before build, with the register attached.", "Deployment survival at team-withdrawal plus 90 days.", "Dependencies that failed versus dependencies anticipated in the register.", "Days from program end to published findings, pass or fail." ], "cases": [ { "verdict": "good", "title": "Off-grid solar as a stack", "summary": "The businesses that displaced kerosene shipped the lamp with distribution, consumer financing and after-sales service. The lamp alone had been available for years." }, { "verdict": "bad", "title": "Shyp and the last-mile economics", "summary": "A courier would come to your door, pack the thing and ship it. Venture-backed, beautifully executed, and shut down once the cost of the pickup refused to fit the price customers would pay. The app was the easy part." }, { "verdict": "bad", "title": "Zume's pizza robots", "summary": "A $2.25 billion valuation, multiple pivots, and a shutdown after years of trying. Automation inserted into a chain whose other steps had not been solved as a system." } ], "reading": [ { "title": "The Wide Lens", "note": "Ron Adner, 2012. Co-innovation and adoption-chain risk. Why good products fail inside bad ecosystems." }, { "title": "Crossing the Chasm", "note": "Geoffrey A. Moore, 1991. The whole product, and why a partial one does not cross." } ], "paper": "https://cei.inc/colab/ei/papers/minimum-viable-ecosystem.html" }, { "n": 9, "id": "own-the-ground", "row": "Ecosystem", "column": "Investment", "pillar": "Patient Capital", "question": "Do you own what the business runs on?", "title": "Own the ground", "summary": "We buy the building, because long-horizon money has to rest on an asset.", "thesis": "Stakeholders in one room for days requires a building. Leasing it makes the method a cost. Owning it makes the method an asset. So PropCo is designed to hold the property and OpCo to run the programs on it, with the property earning on the days no session runs. We have not bought a building yet, which is why this cell is marked the way it is.", "requires": [ { "heading": "What the practice requires", "items": [ "PropCo buys and holds land and buildings, financed on the research horizon rather than a fund's exit clock. It leases to OpCo at market rate.", "OpCo carries staff, labs and programs, and pays rent. Occupancy becomes a line item instead of overhead.", "Non-program days are sold. Offsites, training, retreats, convenings. Utilisation is a tracked metric.", "Long-horizon capital is secured against the property, so the downside does not rest on programming." ] }, { "heading": "The precedent", "items": [ "Harry Sonneborn's insight at McDonald's was that the durable business sat in the land beneath the restaurants.", "McDonald's Franchise Realty Corp. was formed to hold it.", "The separation of property from operations has outlasted every menu decision since.", "The often-quoted line about being in the real estate business rather than the hamburger business comes from the 2016 film The Founder, not the record. What lasted was the structure." ] }, { "heading": "Real, Win, Worth It", "items": [ "This cell also carries the business case, because owning the base and having a base worth owning are the same question asked twice.", "Real. Is the need real, and is there a market that will pay for it?", "Win. Can we win it? Do we have an edge somebody else would have to buy or build?", "Worth It. Is the return worth the risk, on the horizon we have stated?", "A venture that fails Real gets closed. One that fails Win gets a different owner. One that fails Worth It gets a smaller scope or a longer clock." ] }, { "heading": "Where it goes wrong", "items": [ "Long-dated operating obligations against short-dated revenue. Renting the thing you depend on.", "Treating a campus as a brand expense rather than an asset with a utilisation number.", "Collateral that is described but not held." ] } ], "measures": [ "Real, Win, Worth It scored and written down before capital commits.", "Property equity as a share of invested capital.", "OpCo rent coverage ratio.", "Non-program utilisation. Days let, and revenue per available day.", "Debt service coverage at PropCo.", "Program cost with occupancy included, per participant-day." ], "cases": [ { "verdict": "good", "title": "McDonald's Franchise Realty Corp.", "summary": "A decision to hold the land under the operating business. It turned a franchising company into a property company with a restaurant tenant." }, { "verdict": "bad", "title": "Leased, not owned", "summary": "WeWork carried decades of lease obligations against monthly revenue. It held the risk of ownership without ever holding the asset, and the bankruptcy that followed came from a valuation the business itself never earned." }, { "verdict": "ugly", "title": "FTX and the unsaid collateral", "summary": "A $32 billion exchange collapsed when it emerged customer deposits had been lent to a hedge fund controlled by the founder. Assets described to counterparties could not be produced." } ], "reading": [ { "title": "McDonald's: Behind the Arches", "note": "John F. Love, 1986. The franchise-realty structure, and why the land, not the food, built the company." }, { "title": "The Real Estate Game", "note": "William J. Poorvu, 1999. Underwriting discipline for property inside an operating business." } ], "paper": "https://cei.inc/colab/ei/papers/own-the-ground.html" } ] }