Andy Grove ran Intel through the brutal transition from memory chips to microprocessors. In 1983 he wrote down how he actually managed, and the book has outlived almost every other management text because it treats management as an engineering problem, not a personality contest. These are the principles in my own words, and why each one earns its keep in a real business.
Grove's central equation: a manager's output equals the output of their own organization plus the output of the neighboring organizations they influence. You are not paid for the work you personally do. You are paid for what your team and the teams around you produce.
Why it matters in business: the most expensive mistake a new founder-operator makes is staying the best individual contributor. Every hour you spend doing the work yourself is an hour not spent multiplying ten other people. If your output is just your own two hands, you are an employee with a title, not a manager.
Leverage is output divided by the time spent producing it. A small action with large downstream impact is high leverage; busywork is low leverage. Grove flags three repeatable high-leverage moves: gathering information, making decisions that unblock many people, and "nudging" people toward the right path. Training your team and motivating them are among the highest-leverage things you can ever do, because the payoff compounds for years.
Why it matters in business: the art is selecting one, two, or three high-leverage activities and deliberately neglecting the rest. The flip side is negative leverage, just as powerful in the wrong direction: a delayed decision, meddling in work you delegated, abdicating instead of supervising, or an interruption at the wrong moment can wipe out a whole team's week.
Run your function the way a factory runs a line: with a small dashboard of indicators you check daily, not a quarterly report you read after the damage is done. Grove's twist is to pair indicators so they police each other. Measure volume and quality. Measure speed and error rate. A single metric always gets gamed; a paired one cannot be gamed without the partner metric exposing it.
Why it matters in business: "leads booked" with no "leads that showed up" produces a sales team optimizing for garbage. Pair the count with the conversion and the incentive self-corrects.
The earlier in a process you detect a defect, the less it costs to fix. Grove built quality checks into each stage rather than inspecting only at the end. Inspect inputs, monitor the process, and check outputs, and do the catching at the lowest-value stage you can.
Why it matters in business: a wrong assumption caught in a one-page spec costs an afternoon. The same wrong assumption shipped to customers costs a recall, refunds, and reputation. Push your detection upstream where the value at risk is smallest.
There is no single right way to manage a person. The correct style depends on their Task-Relevant Maturity (TRM): how much experience that specific person has with the specific task in front of them. Low TRM needs close, structured, hands-on direction. High TRM needs delegation, communication, and emotional support. The same star employee can be high-TRM on a familiar task and low-TRM the moment you hand them something new.
Why it matters in business: micromanaging a veteran insults them and wastes your time; turning a novice loose with "figure it out" guarantees a fire. Match the leash to the task, not to the person's title or your general opinion of them. See once a week one-on-ones for inexperienced reports and roughly monthly for veterans.
Grove treated the 1:1 as the highest-leverage scheduled meeting a manager has. The cadence scales inversely with TRM: frequent for the inexperienced, sparse for the seasoned. Critically, the meeting is the report's meeting, not yours. Make them explain the situation. The more they articulate the "why" behind their choices, the faster you read their real TRM and the more they own the outcome.
Why it matters in business: a single well-run 1:1 can redirect weeks of a person's work. That is the definition of leverage, and it is nearly free.
Grove adapted Drucker's Management by Objectives into what he called "iMBOs" at Intel around 1971. The system answers two questions: where do I want to go (the objective), and how do I pace myself to know I am getting there (the key results). John Doerr learned it inside Intel, renamed it OKRs, and later carried it to Google, where it became the goal-setting standard for modern technology companies.
Why it matters in business: an ambition with no measurable checkpoints is a wish. OKRs force you to convert "grow revenue" into a number you can fail at, which is the only kind of goal that actually changes behavior.
These principles are distilled in my own words from High Output Management by Andrew S. Grove (Intel CEO, 1983). The book itself is the primary source; the sourced summaries above corroborate the specific figures and the OKR lineage. Read the original, it is short and dense.