Jon McNeill on Tesla’s speed: delete steps before automating

Jon McNeill says Tesla had to grow rapidly without a marketing budget, and that Elon Musk saw speed in decisions, cash flow and response as a competitive advantage. McNeill later describes a five-step method that questions requirements before automating any process.

McNeill’s first in-person meeting with Musk took place at SpaceX. They walked the factory and examined a manufacturing problem rather than staying with general interview questions. Musk told him success would mean arriving at similar answers from the same inputs.

The Model X Falcon Wing doors were difficult to align. McNeill noticed workers had to fasten bolts from underneath without seeing them and suggested a jig to guide the work. He says the team built one overnight, an example of moving quickly once a problem is understood.

Asked about burnout, McNeill says not every minute requires an immediate response. Teams spend time understanding and breaking down problems, then move quickly once they know what to do. He initially hesitated to join a large company after building six businesses of his own.

Before formally joining, he asked Musk about Tesla’s biggest problem and heard demand. About halfway through a quarter, the company had reportedly sold 3,000 of roughly 12,000 cars it had promised to deliver. McNeill offered to investigate for a week.

He visited eight stores, took test drives under different email addresses and received no follow-up. Sales operations told him around 10,000 people had taken a test drive in the preceding 30 days without a callback.

Staff were incentivized to provide test drives, not explicitly to ask for an order. McNeill asked operations to withhold new test-drive leads until prior prospects had been called. He says orders quickly followed, after which he realized he had made a company-wide decision before he was even employed.

He called Musk to apologize and explain. After a long pause, Musk said it was a great decision. McNeill took that as a lesson about the culture: act quickly on first principles rather than wait for a long hierarchy of approvals.

McNeill says Tesla encouraged direct communication with the person who had an answer. As the business grew, his team built supporting structures and offered “battlefield promotions”: people who performed well could take on large new responsibilities before they had held similar roles.

He and Musk experimented with three-sentence emails: state the problem, the proposed solution, and the next step or request for resources. Musk also wanted bad news immediately because waiting would only make it worse.

For disagreements, McNeill recalls advice from Sheryl Sandberg: meet regularly, clear frustrations first, then discuss strategy. He says working with Musk pushed him away from avoiding conflict and toward addressing problems while a team can still learn from them.

Direct feedback should be specific and accompanied by a willingness to help, he says. Rather than leave someone with a vague threat about their job, a leader can explain what failed, show respect for the person and work through what needs to change.

Tesla did not want sales commissions because Musk wanted stores to feel different from traditional car dealerships. McNeill’s team instead displayed a sales leaderboard with a performance line. People worked to get above it even though no commission was attached.

McNeill frames capital allocation as a risk decision: look for asymmetric upside, and avoid downside-heavy bets unless the downside can be addressed. He says that framework helps a team understand why it is committing money or declining an opportunity.

Parrish raises the safety stakes of frequent Autopilot updates. McNeill describes disengagement data being fed into simulations and says corrections were released after repeated simulation. He argues that earlier releases create faster feedback; this is his account of the trade-off, not a demonstrated safety result.

McNeill says not to take foolish bets, but argues against waiting for perfect certainty when a change has been tested and the team is highly confident. He sees software updates as a way to improve a car during ownership rather than treat it as a static product.

His decision checklist starts with the desired outcome and asks what must be true to achieve it. Next estimate the required investment and compare the resulting impact against the effort and competing uses of scarce time and capital.

Opportunity cost is the list of things a team chooses not to do. McNeill says raising more capital can relax a constraint but never removes all constraints, and credits Sandberg with teaching him that the “no” list should be longer and clearer than the “yes” list.

McNeill argues that waiting for perfect information delays learning. Starting with incomplete information can reveal constraints and correct false assumptions, provided the decision maker learns from what happens rather than treating early guesses as facts.

He describes Musk as unusually able to focus on what must be true after a setback. Recalling failed SpaceX launches, McNeill says Musk concentrated on the technical conditions for the next launch rather than remaining focused on the emotional impact of failure.

Product meetings began with Musk asking McNeill what the market wanted. McNeill recalls customers wanting a Model X bench seat for three child seats; Musk and engineering leaders immediately began working through how it might attach to a chassis not designed for it.

McNeill says product teams were expected to show working software or physical systems rather than rely on slides that might hide flaws. Leaders could touch components, inspect wiring and ask engineers questions about the actual design.

Tesla executives selected cars at random from the production line, drove them home and returned notes to engineers the next morning. Choosing an unprepared car was intended to reveal what customers actually received rather than an executive-ready sample.

McNeill also read Reddit and forum feedback while Musk followed Twitter, he says. Direct stores and service centers gave the company information about repairs; using the product and observing customers provided another rapid feedback loop.

He spent roughly one day a week on the front line, visiting stores, service centers or competitors and asking customers what they liked and what would persuade them to buy. He combines those observations with data and design instinct.

McNeill says great products start from simple requirements and design. He contrasts taking orders for a highly sought-after product with actively selling a product against a strong competitor; customer pull and distribution both still matter in the stories he tells.

McNeill recounts a story about Steve Jobs setting a small number of requirements for the iPod: hold 1,000 songs, find a song quickly and make the device beautiful. He presents the story as an example of excluding features that did not support the core goals.

At Tesla, Musk told him the paid marketing budget was zero while the company needed rapid growth. McNeill says the team developed Musk’s social audience, announced product features selectively, generated press coverage and reused that coverage across channels.

Rapid growth still required physical infrastructure: factories, supply chains, distribution and service. McNeill says the team found many repairs could be performed in a customer’s driveway and expanded mobile service rather than building as many service centers.

He describes a manufacturing push to simplify the factory by casting large vehicle structures instead of welding many pieces. He credits Doug Field and the engineering team with pursuing the difficult manufacturing problem; the claimed competitive advantage is his assessment.

McNeill recalls Field rolling a Hot Wheels car across a table to explain the casting concept. Scaling the idea from a toy to a large vehicle part required overcoming pressures, materials and manufacturing constraints that had not yet been solved.

He also recalls a conversation in a car in Hong Kong about the slow speed of typing on a phone. Musk wondered whether people could think directly into a machine and asked staff to find brain-computer-interface researchers; McNeill presents this as an early germ of the Neuralink idea.

The five-step algorithm emerged, McNeill says, from reviewing mistakes. First question every requirement. Second delete as many process steps as possible. Third optimize the simpler process manually; fourth run it faster to expose failures; fifth automate only once it works.

His practical exercise is to map each process step and circle the ones customers actually pay for. The point is to examine why the other steps exist and whether the desired outcome can be delivered with fewer of them.

Speed acts as a stress test: a process run faster exposes quality problems and failure points. McNeill says his teams had made the mistake of automating too early or accepting requirements without questioning them, locking complexity into the work.

As a non-manufacturing example, McNeill describes examining thematic exchange-traded funds and questioning market-cap weighting. He says his firm created funds weighted around companies’ contributions to a product’s bill of materials rather than simply company size.

He gives AI data-center infrastructure as an example, including suppliers of fiber, power electronics, networking and memory. He reports strong past performance for his firm’s funds and attributes it to this approach.

When simplifying, McNeill begins by taking an existing thing apart and asking why its assumptions exist. He credits his grandfather’s repair shop with teaching him that disassembly can reveal how to fix something.

He compares business processes to a garden that accumulates weeds. Musk, he says, repeatedly asked whether an established tool or subscription was necessary and whether the company could approach the work differently.

McNeill says his family prompted him to leave Tesla. During a difficult period, he felt unable to help Musk with personal struggles and did not realize how much the stress had affected his own mood and behavior at home. His family told him they did not like who he was becoming.

He told Musk that he could not provide the help needed and that the situation was affecting him. McNeill describes leaving as giving up a job and colleagues he loved, not as a professional failure; he later continued working with several former teammates.

Asked why some demanding leaders bring out exceptional work, he points to the size of the challenges they give people. A skilled person may discover capabilities they did not know they had when asked to solve a difficult problem.

Parrish asks whether people copied Steve Jobs’s abrasive style after reading his biography. McNeill says leaders need an authentic style rather than imitation. He recalls Walter Isaacson pointing to Satya Nadella as an effective leader with a different manner.

The “fire yourself” exercise came from Calvin McDonald, McNeill says. A leader asks what a successor would immediately change if they were gone, lists those obvious actions, then asks why fear, competing priorities or reluctance to disappoint people has prevented acting.

Discussing Lululemon’s CEO selection, he says the board identified product judgment, talent, scaling and experience with a vertically integrated retail business as important. Different board members assessed candidates from their own perspectives against those shared requirements.

To identify talent, McNeill asks candidates to explain the details of a problem, the solution and their own role. Deep follow-up helps distinguish someone who did the work from someone describing another person’s achievement.

He says he spent about 40% of his time interviewing at Tesla. In a final interview, he first tested the candidate’s depth, then explained the culture, expectations and behaviors that would make the person successful as the company grew.

For trust with a direct report, McNeill asks for concise, factual communication and a clear request. If someone sees a mistake or opportunity, he would rather hear it directly than have to infer a hidden agenda.

He recalls a failure in his second company: he favored organic growth while a co-founder favored acquisitions, and they did not reconcile the disagreement for years. Although the business was sold successfully, employees were caught between competing visions and some felt relief when it ended.

McNeill takes responsibility for not confronting that conflict. He recommends The Goal, The 7 Habits of Highly Effective People and Unreasonable Hospitality as books that influenced his approach to operations, life and treating customers with dignity.

He distinguishes vertical mentors, who have already traveled the path, from horizontal mentors facing similar problems at the same time. Peer groups can examine each other’s businesses from a different angle under confidentiality.

McNeill says his strongest experienced mentors were patient teachers. One, Intuit founder Scott Cook, showed him a product exercise that brings users into a room while the management team watches and listens.

Users list three fixes and three missing capabilities, then each receives five votes to allocate among the suggestions. McNeill says the votes often concentrate on a few priorities, while leaders hear the emotion behind the requests.

He brought the exercise to Lyft, where Woody Hartmann and colleagues ran it with drivers. McNeill says the resulting insights informed the product roadmap for roughly a year.

McNeill says Lyft was positioned as an alternative when customers were unhappy with Uber, and that the team doubled share and revenue during his time. He sees positive differentiation against alternatives as an important lesson.

His second ride-share lesson is to resist competing only on price. When drivers and passengers see services as interchangeable, he argues, the product becomes a commodity and margins collapse. He wishes he had pushed harder for a stronger brand position.

As a board member, he tries to be useful to the CEO while carrying a fiduciary duty. During store visits he asks staff: if you were CEO for a day and could change only one thing, what would it be? He brings the resulting themes back as input, not as unquestionable fact.

DVx screens ideas for a large addressable market validated through customers, gross margins above 50%, a difficult technical problem, favorable competitive dynamics and a path to liquidity. The team prefers not to enter an already crowded category merely because it is fashionable.

The venture builder initially develops the product and tests whether customers will pay. McNeill says approaching $1 million in contracted annual recurring revenue can be a signal to recruit an experienced entrepreneur who brings market knowledge and a team.

For time management, he blocks time for exercise, family and focused work. He tries to handle an email when he first opens it rather than repeatedly returning to it, and reserves long stretches for difficult problems.

He says Musk’s willingness to tackle large technical challenges influenced him. At DVx, the team similarly looks for hard problems that have kept others away, because solving one may create a defensible business.

McNeill says his motivation shifted after an early company outcome, when employees told him their equity helped buy a home or fund a child’s education. He came to value creating durable companies and jobs as much more satisfying than the number in his own bank account.

Asked what success means personally, he says it is having children and a partner who want to spend time with him, time to build companies, and time outside in nature.