The Algorithm

The Algorithm backs three of Tom's lessons. Under each of his lessons, what the book says about it, and on which pages.

• Week one is an audit, and two formats are never up for testing

  Both say to learn the real process before changing anything, and to keep a few things fixed so that effort goes where it matters. pp. 34 and 78. In the book: The Algorithm, Map the process and circle what the customer pays for, Cycle time and touch time, Karim's three conditions.

• Assumptions made too fast, and risk avoided

  McNeill's culture of questioning requirements and backing people who take initiative, even when some fail, matches the idea of backing yourself on data plus gut. pp. 19 and 43. In the book: The Algorithm, Map the process and circle what the customer pays for, Cycle time and touch time, Karim's three conditions.

• Move the people you have before you hire

  Nikki Monterroso's battlefield promotion and the reuse of existing staff in new roles show the same preference for moving people who know the process. p. 52. In the book: The Algorithm, Map the process and circle what the customer pays for, Cycle time and touch time, Karim's three conditions.

McNeill argues that any business can grow exponentially by running every process through five ordered steps: question every requirement, delete every possible step, simplify and optimise, accelerate cycle time, and automate last. Most rules are customs, most steps are not paid for by the customer, and speed only pays once the process is simple. Automation done early freezes a bad design in code. He adds three cultural ingredients that make the steps stick: define the product as the customer's whole experience, have the top executive hold weekly urgent problem-solving sessions with small teams, and use your own product daily. Tesla, GM, Lululemon and McNeill's DVx start-ups serve as case studies.

• Treat rules as recommendations until proved otherwise; ask who made the rule and whether they have data (p. 19).
• Circle the steps the customer pays for and delete the rest, then expect to reinstate about 10 percent (p. 34).
• Cut choices: 300,000 configurations down to two packages lifted sales by more than 20 percent (pp. 36 to 37).
• Replace long training with one aspirational sentence people can act on: 'Be so great that they'll talk about you at dinner' (p. 49).
• Cycle time minus touch time is the prize; a four-week car repair had eight hours of work (p. 61).
• Automate last: hand-build the process first, as Amazon, DoorDash and the Tesla tent line did (pp. 78 to 79).
• Run the first version of a tool by hand with Excel, Forms and Zapier before writing software, as Curbee did over seven months (p. 86).
• A small team reporting to a senior executive who clears roadblocks is the structure for innovating at speed (p. 69).
• Hold a weekly problem-solving cadence with the person who can decide, and know the details (pp. 107 to 111).
• If customers need YouTube to find a switch, the design needs a tweak; use your own product and competitors' (p. 123).

• What does AI actually change about the shape of a content team?

  Stash built an AI coach with a team of seven after a one-month prototype, and the book says to automate only once the human process is understood.

• What is the smallest team that can produce at the volume I am used to?

  GM's Hummer team and Stash's 'Ninjas' show small teams with one decision-maker and a senior sponsor shipping at speed (pp. 67 to 69, 115 to 116).

• How do you price and sell a tool you built for yourself?

  Curbee began as DVx's own mobile repair business used as a lab, run by hand for seven months, and only then became a subscription product for 500 dealerships (pp. 83 to 86).

Tom works with creators and founders one to one, from a single session to a fractional role. Work with Tom