ANALYSIS · Time

Executive Time Management: What CEO Calendar Studies Actually Show

What diary studies of chief executives, from a small Harvard sample to more than a thousand manufacturing bosses, show about where a leader's hours go and which of them can be taken back.

Rendal Post Analysis Desk · Published October 2, 2026

Executive time management is usually sold as a personal productivity problem. Two detailed diary studies of how chief executives actually spend their weeks point somewhere else. In our reading, a CEO's calendar works less like a to-do list and more like a management system: it shows the organization what the leader cares about, and the Harvard study, as we read it, makes the case for deciding in advance where the hours go instead of letting the most insistent demands decide. Below we set out what two diary studies measured, where they differ, and what the owner of a mid-sized company can take from them.

What the Harvard diaries recorded

The Harvard evidence comes from Michael E. Porter, a university professor at Harvard Business School, and Nitin Nohria, then dean of the school, who tracked the time of 27 chief executives and published the results in Harvard Business Review's July-August 2018 issue (MGMA). The participants, all alumni of the school's New CEO Workshop and mostly running large, complex companies, had their time coded by their executive assistants in 15-minute blocks around the clock, with each CEO then checking the coding (MGMA). Fortune reported that the group comprised 25 men and 2 women and that the study ran from 2006 to 2018 (Fortune), while a later academic paper describes the Harvard work as following 27 CEOs for three months (Bandiera, Prat, Hansen and Sadun). In total, more than 60,000 hours of CEO time were coded (MGMA).

The headline averages are demanding: the CEOs averaged 9.7 hours of work per weekday and 62.5 hours per week, and they still averaged 3.9 hours of work on each weekend day and 2.4 hours on each vacation day (Fortune). Face-to-face interaction took 61% of working time, electronic communication 24%, and calls, reading and written correspondence the remaining 15% (MGMA). They averaged 37 meetings a week, which absorbed 72% of their working hours, and only 47% of their work took place at company headquarters (MGMA).

The figure that deserves the most attention, in our reading, is a different one. At the start of each quarter the participants set an agenda with the researchers, and on average just 43% of their time went to activities that advanced it, a share that varied considerably from one CEO to the next (Fortune). Fortune reported that the CEOs who spent the most time on their agendas said they felt better about how they used their time (Fortune). That is a self-reported feeling, not a performance measure.

A larger sample, closer to the mid-sized firm

The Harvard group is small and tilted toward big companies. Economists Oriana Bandiera, Andrea Prat, Stephen Hansen and Raffaella Sadun, in a paper published in the Journal of Political Economy in 2020, took another route: rather than following a few executives for months, their team phoned CEOs or their personal assistants every day for one week and built diaries for 1,114 CEOs of manufacturing firms in Brazil, France, Germany, India, the UK and the US (Bandiera, Prat, Hansen and Sadun).

For an owner, that sample is unusually relevant. The median firm had 300 employees and $35 million in sales, and 57% of the firms were owned by a family (Bandiera, Prat, Hansen and Sadun). The average CEO logged about 50 working hours over the survey week and spent 70% of that time interacting with others, whether in meetings, plant visits, calls, videoconferences or email (Bandiera, Prat, Hansen and Sadun).

The two studies measure the unexpected differently, and the numbers look far apart. Porter and Nohria found that about 36% of CEO time went to handling unfolding issues, internal and external, though pure crisis management averaged only 1% (MGMA). The larger diary study found that activities defined as emergencies took just 4% of CEO time and that just under 10% of planned activities were cancelled (Bandiera, Prat, Hansen and Sadun). In our reading, the gap may reflect what each team counted, since a scheduled meeting can still be a reaction to someone else's problem.

Two ways to fill a week

The economists ran a machine learning algorithm over the diaries and identified two behavioral types: "leaders," who hold high-level meetings spanning several functions, and "managers," who meet individually with core functions (Bandiera, Prat, Hansen and Sadun). In the paper, the leader pattern involves more time with other C-suite executives, while the manager pattern centers on plant visits, production staff and suppliers.

Leader-type CEOs were more likely to run more productive and more profitable firms, a correlation, not a controlled experiment: a one standard deviation increase in the behavior index was associated with roughly 7% higher sales after controlling for labor, capital and other firm-level factors (Bandiera, Prat, Hansen and Sadun). The authors report that the productivity gain after a leader is appointed builds gradually, and their abstract states that "it takes three years for a new CEO to make a difference" (Bandiera, Prat, Hansen and Sadun).

The authors are explicit about the limits, and we repeat them. They write that their results "should not be taken as evidence that all CEOs should behave like leaders," because the behavior that maximizes performance appears to depend on the fit between CEO and firm, and because their data cannot separate what CEOs do from traits the researchers could not observe, such as charisma or vision (Bandiera, Prat, Hansen and Sadun). Their model estimates that 17% of firms ended up with the wrong type of CEO, with such mismatches far more common in lower-income countries, at 36% against 5% (Bandiera, Prat, Hansen and Sadun).

Where the recoverable hours sit

Porter and Nohria's recommendations, as reported by MGMA, focus on a few leaks. Routine duties such as review meetings, board meetings and earnings calls took 11% of CEO time on average, and the authors wrote that some CEOs, "especially those who had been [chief operating officers]," overinvested in reviews their direct reports could have handled (MGMA).

Meetings are the largest block. The authors suggested questioning the default one-hour slot in favor of 30 or even 15 minutes, and MGMA quotes one CEO's rule for meeting requests: "whatever they ask for, cut it in half" (MGMA). Small formats already dominated: one-on-one sessions made up 42% of meetings and groups of two to five people another 21% (MGMA). Email is the third leak. The authors recommended having an executive assistant filter incoming messages, and they wrote that there is "no substitute for being disciplined about resisting the siren call of electronic communications" (MGMA).

At the same time, Porter and Nohria argued against a calendar that is booked solid. The share of time CEOs left for spontaneous meetings ranged from 3% to 61% across the group, and the authors warned that leaders whose assistants act as gatekeepers and turn too many people away "risk being viewed as imperious, self-important or out of touch" (MGMA).

Why presence still counts

Jeffrey Sonnenfeld, senior associate dean for leadership studies at the Yale School of Management, draws on the Harvard findings to argue that virtual meetings are necessary for leaders of sprawling enterprises but no substitute for engaging in person (Yale Insights). Citing Porter and Nohria, he puts the CEO workday at roughly 10 hours, more than 61% of it spent face to face, and relays their view that in-person exchanges are where CEOs exert influence most effectively and learn what is really happening (Yale Insights).

He quotes Anne Mulcahy, the turnaround CEO of Xerox, who told him "It is important to have good sources" and said face-to-face communication with all constituencies took a big piece of her time, with travel of more than 100,000 miles a year (Yale Insights). Presence can also backfire. Sonnenfeld recounts that Graham Phillips, on taking over as CEO of Ogilvy & Mather, told his top leaders around the world, "If you see me, that means you’re in trouble," an approach Sonnenfeld judges hardly inspiring (Yale Insights).

What to do with this

For the owner of a mid-sized company, the evidence in our reading supports a few concrete moves, not a new productivity method. Start by writing down the agenda only you can advance and comparing your actual calendar against it every quarter, the review Porter and Nohria urged on their participants; their warning was that without such an agenda, "demands from the loudest constituencies will take over" (MGMA).

Then audit the routine. Reviews a direct report can run, default meeting lengths and the inbox are the areas the Harvard authors singled out (MGMA). Keep part of the week open for unplanned conversations, since those authors tie accessibility to a CEO's legitimacy.

Finally, treat the leader-versus-manager distinction as a question of fit, not a verdict. The diary study links broad, cross-functional meetings with senior executives to better performance on average, but its authors caution that the right behavior depends on the firm (Bandiera, Prat, Hansen and Sadun). In our reading, a founder whose plant still needs daily attention may be the right kind of CEO for that company today, and the useful test is whether that pattern is a deliberate choice, and whether the team below is strong enough to take on more. On that point, the Harvard CEOs' top regret, according to MGMA, "was not setting high-enough standards in selecting direct reports" (MGMA).