The Algorithm

The problem it deals with

Most businesses add steps over time and almost never remove them. A rule gets made for a good reason, then becomes a policy, then a requirement nobody questions. Training grows to a month, approvals pile up and software is bolted onto a process that nobody has examined. McNeill, who was president of Tesla from 2015 to 2018, offers a short sequence for undoing this, the one Elon Musk called the Algorithm.

What changes after you read it

You start asking of every step whether the customer pays for it. McNeill suggests mapping the process on sticky notes, circling the steps the customer values and trying to remove the rest. You also begin to separate cycle time from touch time. His example is a car repair that took more than four weeks and only a few hours of actual work. The gap between the two is where the improvement sits.

Why it suits an owner of a growing business

Growth tends to bring more meetings, more checks and more handoffs. The book argues that you can grow faster by taking steps out instead of adding people. It also gives practical advice for a small company: set an outsized goal with a deadline, put a small team on it, and meet every week to hear what has changed. Those habits cost nothing.

The order matters

The most useful idea for me is automate last. McNeill describes Tesla's attempt to run an almost fully automated line for the Model 3, which did not work, and the move to hand assembly in a tent so that the team could learn the real steps first. His later software start-up did the same: it ran its bookings from spreadsheets and forms for months before writing any code. If you plan to hand work to software or AI agents, read that chapter first.

When to read it, and when to skip it

Read it when a process feels slow and you cannot say why, or before you automate something. Skip it if you want a deep treatment of lean manufacturing, because the book is a set of stories with short guidance sections, not a method with tools and templates. It is also written from the author's own vantage point and draws heavily on his own companies and boards, so it reads partly as a showcase. He is open that Musk has blind spots and that his own time at Tesla ended in 2018.

How it connects to decisions you can log

The weekly cadence is a good fit for a decision log. Each week the owner asks the small team for what has been tried, what was learned and what the result was on the two or three biggest problems. Recording those answers builds a history of which steps you removed and why. Once a process is simple and written down, it is ready to hand to a person or an agent as a repeatable playbook.

McNeill argues that a short set of principles can make any business grow quickly, whatever its size or industry. He learned them from Elon Musk while serving as president of Tesla, and he has used them since at other companies, including lululemon and General Motors, where he sits on the boards. Part I covers the five steps of the Algorithm, and Part II covers three cultural practices that make the steps stick. He says the book is not about Musk's politics, and that much of the genius at Tesla and SpaceX comes from engineers and technicians who were free to question everything.

Introduction

McNeill tells how he met Musk and fixed a stalled sales process by walking through it himself. He took test drives at eight Tesla stores and never got a follow-up, found about five thousand similar leads and traced the cause to commissions that rewarded leads and not sales. Sales rose within days. He then ran operations for thirty months, during which revenue grew from about two billion to about twenty billion dollars. He explains why he left in 2018 and says Walter Isaacson encouraged him to write the book.

Step 1: Question Every Requirement

Rules become requirements over time, and many turn out to be recommendations, customs or a supplier's habit. The chapter opens with the fourteen months of negotiation that won Tesla a wholly owned plant in China. It continues with the casting idea that removed the body shop: a chassis built from three parts instead of three hundred. Other cases include a way to identify a driver's phone by their movements, and a cyber-insurance start-up born from a seven million dollar premium that looked too high for a retailer. The aim is to treat every no as a possible yes, while remembering that physics is not negotiable.

Step 2: Delete Every Possible Step in a Process

Musk set a goal of growing digital sales twentyfold. McNeill's method is to ask managers to list every step, circle those the customer does not pay for and delete most of them, accepting that perhaps a tenth will need to come back. Counting clicks led to fewer buyer types and fewer options, a purchase in twelve clicks and a large rise in online sales. The chapter also covers the Hurricane Irma software update that gave Florida drivers extra range and a restaurant app that removes the bill. It suggests describing your process to an ignorant robot to expose steps you can drop.

Step 3: Simplify and Optimise

Simplify only after deleting. A month of sales training became a single aim, to be so great that customers talk about you at dinner. At delivery centres a team replaced a long, fearful pre-delivery routine with a clear standard, a protective cover that cost under fifteen dollars and a fixed sequence of cleaning steps, cutting preparation from hours to minutes. A three-star restaurant kitchen is used to show that the more complex the operation, the more it depends on simple, repeatable steps. The guidance is to map the process, remove steps the customer does not pay for, then practise the rest until it is fast.

Step 4: Accelerate Cycle Time

Speed comes after the first three steps, because there is no point speeding up a process you should have deleted. McNeill defines cycle time, from drop-off to pick-up, and touch time, the hours spent actually working, and uses his own body-shop repair to show the gap. A new planning system gave Tesla real-time visibility across supply chain and factory. At General Motors, an electric truck cut new model development from years to about a year, and at lululemon a rushed Olympic kit and faster delivery followed. He advises aiming for large weekly gains in speed.

Step 5: Automate Last

The Model 3 production crisis shows the risk. The heavily automated line failed, cash was short, and a tent line assembled by hand taught the team the steps and raised output tenfold in six weeks. Parking-lot inventory was tracked with paint pens and cones before it became the Unicorn Report dashboard. A mobile-repair start-up ran on spreadsheets, forms and maps for seven months before it built software, using Zapier along the way. The advice is to hold back the coders, because code built too early is expensive to change.

Radical Rethinking

Apply all five steps and the result is often a new way to do the work. Faced with a hundred new service centres Tesla could not afford, a manager redesigned service so that a senior technician met the customer at the kerb and fixed most cars in about twenty minutes, and mobile repair vans followed. The chapter closes with an investment fund start-up that began from the discovery that an obvious product did not exist. McNeill advises surrounding yourself with simplifiers.

Expand the Definition of Your Product to the Customer's Entire Experience

The first cultural ingredient is to take the customer's view. For Tesla the product included charging, financing, service and software, and each pain point was an opportunity. Insurance sold with the car, built on driving data, came to represent a large share of cash flow, and subscriptions for features followed. The chapter also describes Walmart's move into groceries, pharmacy and fuel, and the guidance is to ask what customers do in and around your product.

Inject Urgency and Accountability into Your Organisation

The second ingredient is the weekly cadence. Musk spent a day each week at the Palo Alto engineering centre, chose two or three pressing problems and expected teams to report progress, with solutions and not status. His grasp of the details kept the sessions honest. McNeill now meets his own company heads weekly, and his venture studio runs stage gates with scorecards and drops ventures that miss their targets. The guidance is to name the biggest problem, pick the right small group and meet again in a week.

Eat Your Own Dog Food

The third ingredient is to use your own product as customers do. The author notes that only half the finance executives in one audience used their own company's app. Leaders at lululemon walked stores in secret or tested running gear daily. Asking for unvarnished feedback produced quick wins, such as a weekly ranking email that lifted store sales. Doing this creates a short feedback loop, and teams soon start doing it too.

What's Next

The closing section places the Algorithm in history: Ford's Model T followed it almost step for step, and General Motors answered with differentiation. McNeill says the process never ends, since last year's gains become this year's norms, and he treats it as a survival strategy in a period shaped by AI and electrification.

What to do with it

• Map one process on sticky notes, circle the steps the customer pays for and try to remove half of the rest.
• Write down the requirements behind it and find out who made each one and whether it is a rule, a custom or a preference.
• Measure cycle time and touch time for one job and look at where the waiting happens.
• Run the improved process by hand or with simple tools for a few weeks before you write any software.
• Start a weekly meeting with a small group on your single biggest problem, and use your own product the way a customer does.