Tesla: The Battery Bet That Rewrote the Auto Industry

An Interactive Case Study on Vertical Integration, Dynamic Capabilities, and the Courage to Integrate When Finance Says Outsource

Fremont, California, 2008

You are Elon Musk, and your company has nine million dollars in the bank. The Roadster — Tesla's first car, a bespoke electric sports car built on a Lotus Elise chassis — has just started delivering. Each unit costs more to manufacture than it sells for. Lehman Brothers collapsed six weeks ago. The capital markets are closed to venture-backed auto startups. Your co-founder Martin Eberhard is gone; so is every CEO Tesla has had in its five-year history. The board has just asked you, reluctantly, to take the title. You have three months of runway.

Every adviser in Detroit, every banker on Sand Hill Road, every engineer trained at GM or Ford tells you the same thing: batteries are a commodity input. Buy cells from Panasonic or LG. Don't tie up capital in factories you can't run. Build the car. Let the specialists build the cells. This is orthodox automotive wisdom — 100 years of accumulated Make-or-Buy logic. It is also, you suspect, completely wrong. You will have 12 years and four major decisions to test that suspicion. Get them right and Tesla becomes the most valuable car company on Earth. Get them wrong and you join DeLorean, Tucker, and Fisker in the graveyard of electric dreams.

Cash (Oct 2008)

~$9M

Roadster unit margin

Negative

Employees

~500

Years of runway

0.25

The Protagonist

Elon Musk is 37. He sold PayPal for $1.5B in 2002, put $100M into SpaceX, $70M into Tesla, and $10M into SolarCity. By late 2008 essentially all of that is gone — Tesla has consumed his entire liquid net worth. He is borrowing money from friends to pay rent. His biographer would later write that Musk considered this the worst year of his life. And yet, the memo he is writing to the Tesla board that November is not about survival; it is about what the company should build. He believes — against every piece of auto-industry advice he has received — that the battery cell is not an input to the car. It is the car. Everything else is a container for the battery. And if that is true, then the most expensive, most strategic, most rapidly-evolving component in the vehicle must not be outsourced. This is the belief that will guide every major decision for the next 12 years.

[ Image placeholder — Elon Musk inspecting a battery pack at the Tesla Fremont factory, 2010 ]

The moment the auto industry's oldest assumption was about to be tested

The Aha Moment

The single sentence that explains Tesla

The auto industry treated the battery like a tire. Tesla treated the battery like the engine. Every competitive moat Tesla built downstream — Supercharger network, software, manufacturing scale, vertical integration into lithium refining — flows from that single reframing. Everyone else was selling cars that contained batteries. Tesla was selling batteries that contained cars.

The Challenge

Four decisions separate Tesla the niche EV boutique from Tesla the most vertically integrated car company since Ford's River Rouge. Each is worth up to 25 points. One of them — the Gigafactory bet in 2014 — is arguably the most important capital allocation decision of the 21st-century auto industry. Score up to 100 and you will have seen, in half an hour, the playbook that MBA programs now teach as the canonical case of dynamic capability outperforming financial optimization.

Before you start

What should we call you during the simulation?

Your name appears in the feedback and on the final certificate. We save it to your account so we don’t have to ask again.

Strategic Decision Simulation

Score: 0/100

Decision 1 of 4

Decision 1: The Model S — Where Do You Enter the Market?

It is 2009. The Roadster has proven that electric is not a golf-cart fantasy — it is a sub-four-second sports car. You have secured a $465M Department of Energy loan to build your next vehicle. But what is the next vehicle? The board is split. The Silicon Valley half says go for volume — an affordable sedan at $35k competes with Camry, teaches the mass market that electric is normal, reaches millions. The Detroit hires say at $35k you cannot recover battery cost; you will bleed indefinitely. Your own engineering team says the cells we can buy from Panasonic today cost roughly $1,000/kWh at pack level. To build a 300-mile range car, we need an 85 kWh pack. That's $85,000 in battery cost alone — before the aluminum, the motors, the software, the seats, or the factory. The math is unforgiving.

Cell cost 2009

~$1,000/kWh

Target range

300 mi

85 kWh pack cost

~$85k

DOE loan

$465M

Where do you position the Model S?

Decision 2: The Battery Question — Outsource, Partner, or Build?

It is September 2014. The Model S has shipped 31,000 units in its first full year. Demand is outrunning supply — because Panasonic, your cell supplier, cannot scale fast enough. By 2018 Tesla will need more lithium-ion cells than the entire global market produced in 2013. There are four ways to solve this. Three of them are what the auto industry would do. One of them is what Tesla will do. Panasonic's CEO has been explicit: "We will sell you every cell we can make. We will not build a new factory just for you." That is the constraint. This is, by every credible estimate, the most consequential capital-allocation call Tesla will ever face. A wrong answer here doesn't just slow growth — it caps it.

2013 Tesla cells

~2 GWh

2020 Tesla need

~35 GWh

Gigafactory capex

$5B

Tesla 2014 revenue

$3.2B

How do you solve the battery constraint?

Decision 3: Production Hell — The Model 3 Is Breaking the Company

It is Q2 2018. The Model 3 was supposed to produce 5,000 units/week by December 2017. You are at ~2,000. Cash is burning at roughly $1B per quarter. Short interest in TSLA is the highest of any stock on the NYSE. Wall Street is openly predicting bankruptcy. You have been sleeping on a couch at the Fremont factory for six weeks. Your own general-assembly line, which you over-automated in a bet that "the machine that builds the machine" would be Tesla's real product, is failing at station after station. Robots are handling wiring harnesses that human hands do better. You have 90 days before the DOE loan covenants and your working capital both run out. Four options have been whiteboarded by your engineering and finance leads.

Target units/week

5,000

Actual units/week

~2,000

Cash burn rate

$1B/qtr

Short interest

~25%

What do you do in the next 90 days?

Decision 4: Battery Day — What Do You Do With the Lead?

It is September 2020. Tesla is profitable for the first full year in its history. TSLA is up 700% in twelve months. The Gigafactory Nevada has ramped to ~35 GWh of annual cell output — larger than every other EV maker combined. You have a decisive lead: roughly six years of battery manufacturing learning curve that competitors cannot replicate by writing a check. Now you face a strategic choice about what to do with that lead. Engineering has prototyped a new cell — the 4680 — that is 5× larger, 6× more powerful, and 16% cheaper per kWh. Your Battery Day presentation is scheduled for September 22. What do you announce?

TSLA 12-mo return

+700%

Cell capacity lead

~6 yrs

4680 cost/kWh

-16%

Tesla market cap

$400B

How do you use the battery advantage?

Your Tesla, Scored

Final Score

0 /100

—

The single sentence of this case

When the most strategic input is changing fastest, vertical integration is not a cost — it is the option on controlling your own future.

Decision-by-Decision Comparison

Compare your call with what Musk actually did — and what the counterfactual would have cost.

Frameworks in Play

1) Fine & Whitney — The Vertical Integration Decision (MIT, 1996)

Charles Fine and Daniel Whitney argued that the standard Make-or-Buy logic (integrate if cheaper, outsource if cheaper) ignores two deeper variables: whether the component is strategic and whether its technology is stable or changing. Their 2×2 matrix is the most underrated framework in operations strategy.

	Commodity component	Strategic component
Mature technology	Buy on price (seats, tires, bolts)	Buy with long-term contract (legacy ECU)
Evolving technology	Buy but monitor (emerging sensors)	INTEGRATE — own the capability curve (EV battery 2014, AI compute 2022)

Detroit applied the mature-commodity rule to EV batteries because it was applying an ICE-engine mental model. But lithium-ion cells in 2014 were in the strategic + evolving quadrant: 40% of vehicle cost, energy density improving 7% per year, cost falling 15% per year, with multiple chemistry paths still unresolved. Fine/Whitney says: integrate. Musk did. GM didn't. That single decision, applied to that one cell in that one year, is most of Tesla's moat.

2) Teece — Dynamic Capabilities (Berkeley, 1997 / 2007)

David Teece's question: why do some incumbents survive disruption and others don't, even when they see it coming? His answer: three capabilities, built over time, that cannot be bought or licensed.

Sensing — reading weak signals in technology and markets. Musk tracked lithium-ion density curves from 2003 onward. By 2010 he had a mental model of when ICE parity would happen (~2022) that was wrong by only three years. GM did not.
Seizing — committing resources before the market confirms the thesis. The Gigafactory decision in 2014 could not be justified by DCF — the numbers required a forecast of 2020 demand that no analyst model supported. Musk seized anyway. Seizing is an act of conviction, not calculation.
Transforming — continuously reconfiguring the asset base. Gigafactory 1 (cells) → Gigafactory Shanghai → Berlin → Texas → 4680 cells → cathode refining → lithium rights → Supercharger network. Each step deepens the integration chain. This is why competitors cannot catch up by copying one asset: they would have to copy the transformation sequence.

Fine & Whitney tells you when to integrate. Teece tells you why most companies won't, even when they should. Dynamic capabilities are organizational muscle. You cannot hire for them; you have to build them over a decade of seizing imperfect information.

3) Why this pair and not a third framework

This case is deliberately not about Christensen's Innovator's Dilemma (that is Netflix). It is not about leadership charisma (Musk's visionary style is narrative texture, not the mechanism). It is about the most underrated discipline in industrial strategy: when your most critical input is in technological flux, orthodoxy is the trap. Fine/Whitney gives you the rule. Teece gives you the courage to apply the rule when the CFO's model says no.

What happened next

The Model S shipped in June 2012 at starting prices of $57,400 — the premium-first bet worked. The Gigafactory Nevada broke ground in 2014 and shipped its first cells in 2016. Production Hell ended in June 2018 when Tesla hit 5,000 Model 3/week; Musk had literally slept in the factory for weeks and built a temporary assembly line in a tent in the parking lot (GA4). Tesla posted its first profitable full year in 2020. Battery Day introduced the 4680 cell, internal cell production, and the acquisition of lithium rights in Nevada — the deepest vertical integration in the auto industry since Ford's River Rouge in 1928. By 2023, Tesla's cell manufacturing lead was so large that Ford, GM, and Stellantis were paying Tesla to use the NACS charging standard. Panasonic, BYD, and CATL remained the largest cell makers, but Tesla was the only car company that was also a top-five cell manufacturer. The thesis — the battery is the engine, not the tire — had become, after 15 years and $50B of committed capex, the consensus view of the entire auto industry. The companies that realized it too late (Ford, Toyota, Volkswagen) are now spending tens of billions to catch up on a capability that Musk built with $5B when nobody was watching.