Greg McKeown's essentialism framework—do less, better—applies directly to AI project sprawl. Learn how to identify and kill low-impact AI initiatives to protect team focus and resources.
By OperatorRadar Editorial
Greg McKeown's *Essentialism: The Disciplined Pursuit of Less* (2014) argues that the core problem of modern work is not doing too little, but doing too much. McKeown defines essentialism as the disciplined pursuit of the vital few over the trivial many. The framework rests on three core principles: (1) explore and evaluate options ruthlessly, (2) eliminate non-essentials without apology, (3) execute the essential with precision. For operators, this means saying no to projects that don't align with core business outcomes, even when they seem valuable in isolation.
McKeown wrote *Essentialism* during the rise of productivity culture and the myth that "having it all" was achievable through better time management. The book directly challenged the assumption that more features, more projects, and more optimization always yield better results. It resonated with leaders drowning in competing priorities and decision fatigue. The framework gained particular traction in tech, where feature creep and scope expansion are endemic.
McKeown's core insight: the cost of saying yes to a mediocre project is not the time spent on it—it's the opportunity cost of *not* doing something essential. He emphasizes that elimination is harder than optimization because it requires explicit trade-offs and the courage to disappoint stakeholders. Essentialism is not about doing less work; it's about doing the work that matters most, with full focus and resources.
AI has multiplied the number of plausible projects operators face. Every team now has a backlog of AI experiments: chatbots, automation, analytics, content generation, process optimization. The barrier to entry is lower (off-the-shelf models, no-code platforms), so the temptation to start projects is higher. The risk: teams spread resources across 8–12 AI pilots, none of which reach production or deliver measurable ROI. McKeown's framework becomes more urgent, not less, because the cost of diffusion is now measured in months of engineering time and opportunity to build competitive advantage.
The core principle holds: focus beats breadth. An operator who commits fully to one AI project that solves a $2M annual problem will outperform a team running five AI experiments that each address $200K problems but ship nothing. The elimination principle also remains true: saying no to a plausible AI project (e.g., "we could use AI to improve our reporting") is harder than saying no to an obviously bad one, but it's essential. McKeown's insight that elimination requires explicit trade-off conversations is directly applicable—operators must articulate what *won't* get done if a new AI project is greenlit.
McKeown's framework assumes a relatively stable set of options and priorities. AI changes the landscape faster than traditional business initiatives. A project that was non-essential six months ago (e.g., fine-tuning a model for your domain) may become essential today due to competitive pressure or new capability availability. Operators need to revisit their essential-vs.-trivial categorization quarterly, not annually. Also, McKeown's framework emphasizes individual and team focus; it doesn't address the organizational coordination required when multiple teams are running AI projects that share infrastructure, data, or models.
Use essentialism to audit your current AI project portfolio. For each active or proposed AI initiative, answer: (1) Does this project directly support one of our top 3 business outcomes? (2) If we don't do this, what revenue, cost, or risk impact occurs in the next 12 months? (3) What would we *not* do if we commit full resources to this? (4) Do we have the right team and infrastructure to ship this in 90 days? Projects that fail questions 1–2 should be killed or deferred. Projects that pass 1–2 but require trade-offs in questions 3–4 should trigger explicit leadership conversations, not silent resource drain. The goal: move from a portfolio of 8 pilots to 2–3 focused bets that ship and compound.
Take your current AI project list. For each project, write down the single business outcome it supports and the dollar value of that outcome over 12 months. If you can't articulate the outcome or its value in one sentence, that project is a candidate for elimination. Which projects disappear when you apply this filter?
Greg McKeown
Essentialism: The Disciplined Pursuit of Less
“McKeown defines essentialism as the disciplined pursuit of the vital few over the trivial many, emphasizing that the cost of saying yes to a mediocre project is the opportunity cost of not doing something essential.”
Greg McKeown
Essentialism: The Disciplined Pursuit of Less (Book)
“The book argues that the core problem of modern work is not doing too little, but doing too much, and that elimination of non-essentials is harder than optimization because it requires explicit trade-offs.”
Operators managing AI portfolios often struggle to kill projects because the sunk cost feels real and the upside of "just one more experiment" feels safe. If you're running more than three AI initiatives in parallel, you're likely diffusing resources. OperatorRadar can help you audit your portfolio, quantify trade-offs, and build the case for elimination. [Contact us to discuss your AI project roadmap.]
Request implementationNeed a custom implementation?
Have Ekofi Lyrae design and implement AI agents and automations.