Rethinking Productivity: Why Founders Must Manage Energy, Not Just Time
Most founders traditionally measure their productivity by hours worked—hours spent in meetings, hours shielded for deep work, or simply the total hours logged each day. This made sense when the main bottleneck was completing tasks manually. However, by 2026, the landscape has shifted dramatically. Advanced AI tools and agents can now draft documents, conduct research, write code, summarize information, and follow up on tasks faster than any human team could. Consequently, work no longer waits on your hands—it waits on your judgment.
Every AI agent finishing a task returns something for you to approve, reject, tweak, or redirect. Each new tool adds streams of outputs to read and more decisions to make. Time is no longer the scarcest resource; it is decision energy—the limited supply of clear thinking you bring to the choices that actually move your business forward. Protecting this resource requires more than another productivity app. It demands an operating system for your energy: a set of rules deciding what reaches you, when and how you handle it, how you replenish your mental reserves, and what happens when you run low.
Why Your Energy Is the New Bottleneck
Recent research from the AI rollout era offers consistent insights. An eight-month study at a 200-person U.S. tech company, conducted by UC Berkeley Haas researchers Aruna Ranganathan and Xingqi Maggie Ye, and featured in Harvard Business Review, found that generative AI did not free up employees’ time. Instead, workers completed tasks faster, took on broader responsibilities, and extended their work into more hours, often voluntarily. AI prompts intruded into lunch breaks and evenings, and some workers ran multiple AI agents simultaneously, even during meetings. The researchers warned that constant task-switching and reduced recovery can impair judgment and increase errors.
Supporting this, Boston Consulting Group surveyed 1,488 full-time U.S. workers and reported in Fortune that employees requiring heavy AI oversight expended 14% more mental effort and experienced 12% greater mental fatigue and 19% more information overload. Productivity increased with up to three AI tools but declined when juggling four or more. Lead author Julie Bedard described this phenomenon as “AI brain fry,” with workers hitting their cognitive limits due to excessive decision-making demands.
Even before AI agents, the relentless stream of communications was taxing. Microsoft’s Work Trend Index revealed that heavily pinged Microsoft 365 users receive a meeting, email, or chat roughly every two minutes during core hours. Nearly half of employees and over half of leaders reported feeling that their workdays were chaotic and fragmented.
It is important to note that the classic concept of “decision fatigue,” likened to a depleting battery of willpower, is debated. A large 2025 field study of healthcare professionals found no evidence of decision fatigue in repetitive triage decisions. Nonetheless, the newer AI research confirms a simpler reality: when too many outputs, interruptions, and judgment calls accumulate within the same hours, cognitive fog, mistakes, and loss of focus inevitably follow. Founders, who sit at the apex of these demands, feel this strain most acutely.
Energy, Not Time, Is the Real Budget
The idea that energy—not just time—must be managed is not new. In 2007, Tony Schwartz and Catherine McCarthy argued in Harvard Business Review that while time is finite, energy can be systematically renewed, and leaders should focus on managing energy to sustain high performance. What has changed is the scale of demand. Today, a single founder managing multiple AI agents faces a flood of reviews, approvals, and judgment calls once distributed across entire teams. Schwartz’s advice to work in cycles of effort and renewal remains one of the clearest frameworks for maintaining sustainable productivity.
Think of your energy operating system like a computer’s OS: it doesn’t do the work itself but decides which processes get resources, which wait, which shut down, and how the system recovers when overloaded. Building this system requires attention to four key layers.
Layer One: Reduce What Reaches You
The easiest decision is the one you never have to make. Founders often expend mental energy on hundreds of trivial decisions that should never have reached them—whether to tweak an email, assess an AI summary, or reply immediately to a Slack thread. While each seems minor, collectively they sap the clear thinking needed for strategic priorities like pricing, hiring, and product development.
Begin by setting defaults: write down recurring decisions and turn them into standing rules. For example, automatically approve refunds below a set amount, decline meeting requests without agendas, or send content drafts that follow guidelines without review. Each default decision is one less cognitive load.
Next, limit your tools. The BCG study’s finding that productivity dips after three AI tools is a practical guideline. Audit the AI tools and agents demanding your attention and consolidate or cut back if you oversee more than three. Ask whether each tool removes decisions from your plate or adds to them.
Finally, build approval gates instead of monitoring everything in real-time. Agents should batch outputs and present only items requiring your judgment, accompanied by clear recommendations (“Here are three options; I recommend the second. Approve or change.”). This approach costs seconds rather than an afternoon. Anything below a defined threshold of risk or impact should proceed without your input. The goal is to conserve your attention for decisions where errors would be costly.
Layer Two: Schedule Your Peak Energy
After reducing volume, decide when to allocate your best thinking. Many founders know their peak clarity times—whether early morning, late morning, or evening. The mistake is letting incoming tasks hijack this window. Microsoft’s telemetry shows 11 a.m., a common productivity peak, is also the busiest hour for messages and meetings.
Protect this peak block for decisions only you can make—those shaping the next quarter—such as strategy, key hires, pricing, partnerships, and creative direction. Reserve AI reviews, approvals, and inbox triage for scheduled batches later in the day. The Berkeley researchers recommend batching AI-assisted work and building pauses before critical judgments instead of letting prompts and reviews fragment focus.
It also helps to distinguish between motion and progress. Answering AI-generated requests and clearing notifications can feel productive but often moves nothing meaningful forward. Recognizing this trap of fake work is essential to protecting your highest-value hours.
Layer Three: Refill as an Operating Function
Recovery is often framed as wellness, causing founders to skip it when busy. Instead, treat refilling energy as an operational necessity. Like a machine overheating without cooling, poor recovery throttles judgment and performance. Refilling is maintenance, not a reward.
Sleep is the foundation. You cannot out-schedule chronic sleep deprivation, and successful founders treat sleep as a critical business asset, not a luxury to sacrifice when deadlines loom. Beyond sleep, build short recovery breaks into your day. Many workers in the BCG study reported mental “fog” or “buzzing” that only cleared by stepping away from screens. A walk between peak blocks and review sessions, lunch away from devices, and firm stopping times are essential cooling mechanisms.
Movement deserves equal priority. Schedule it with the same importance as board meetings—if your energy fails, every other commitment suffers.
Layer Four: Throttle When You Are Overdrawn
Every operating system needs a safe mode. There will be weeks when launches, fundraises, or crises drain you faster than you can replenish. The worst mistake is to keep making major decisions at full speed while running on fumes since that leads to costly errors.
Identify your warning signs in advance: rereading the same paragraph multiple times, second-guessing simple decisions, snapping at your team, or compulsively checking your phone when trying to focus. When these arise, activate safe mode: postpone major irreversible decisions until rested, tighten approval thresholds to reduce input, pause non-critical AI experiments, and inform your team of reduced capacity so only urgent matters reach you.
This is not a sign of weakness but of risk management. Like a pilot who respects an aircraft’s limits and flies safely rather than recklessly, you preserve your capacity for the long haul.
Putting Your Energy OS Into Practice
Implementing all four layers at once is unnecessary. Start with a one-week audit: for five days, log every decision reaching you, estimate its energy cost, and note whether it could be handled by someone or something else. At week’s end, categorize decisions into three groups: those that should become defaults, those to delegate to a person or agent with clear boundaries, and those that genuinely require your attention.
From there, establish three to five defaults, cap the number of demanding tools, block your peak decision-making window for critical tasks, schedule recovery time, and define your safe-mode triggers. Review and adjust your energy operating system monthly, aligning it to evolving business needs and tools.
AI offers founders unprecedented leverage, but that advantage only pays off if the person at the center can think clearly. Your calendar shows how you spend hours; your energy operating system determines whether those hours are spent well. In the years ahead, this subtle edge will distinguish founders who scale sustainably from those who burn bright and fast before burning out.
