TL;DR. A week has 168 hours, and the 2025 American Time Use Survey shows how employed adults actually spend them. Converted to weekly totals by CEOtudent, employed adults with no children under 18 log about 61.8 hours in the survey’s sleeping category (which includes naps and sleeplessness), 38.1 hours on work and work-related activities, and 31.7 hours on leisure and sports. With a child under 6, care rises to 13.4 hours a week and leisure falls to 22.3. Inside those totals, work itself has become harder to schedule: Microsoft’s own telemetry (company-reported) finds that 57% of meetings are ad hoc calls without a calendar invite. And the best-known planning tool, the if-then plan, works less well than its famous figure suggests: d = 0.65 across 94 tests in 2006, but d = 0.36 across 642 tests in 2025, or 0.15 after the strictest correction for publication bias. The ideal week that survives this is not a perfect calendar. It is a small set of fixed anchors, a held-back capacity buffer, and a weekly re-plan. The CEO move: decide which hours are not up for negotiation before anyone else claims them. The student move: log what each week actually cost, and use it to correct your next estimate.
This piece is the weekly layer of a larger system. The case for designing systems rather than chasing targets is in Life Design vs. Goal Setting. The step-by-step review that feeds the re-plan below is in The Weekly Review Protocol. Neither of those is repeated here. This article covers the week itself: where the hours go, what to fix in place, and what to leave open.
Why does the colour-coded ideal week break so fast?
The classic “ideal week” is a template calendar: every hour has a colour, every block has a label, and the plan assumes next Tuesday will look like last Tuesday. It fails for three documented reasons.
First, people plan short, and they plan short even when warned. In the original planning-fallacy study, Buehler, Griffin and Ross asked honours students when they would finish their theses. The average best estimate was 33.9 days; the average actual completion time was 55.5 days. Only 29.7% finished by their best-estimate date. Even the estimate made “assuming everything went as poorly as it possibly could” averaged 48.6 days, still short of the actual, and fewer than half (48.7%) beat it. By CEOtudent’s arithmetic, actual time ran 1.64 times the best estimate, and 38.9% of the time the work really took was missing from the plan. A calendar filled to 100% with best estimates is, by construction, a calendar that will overflow.
Second, the work itself arrives unscheduled. Microsoft’s June 2025 Work Trend Index special report, which is company-reported and based on Microsoft 365 telemetry, says employees are interrupted every two minutes during core hours (275 times a day) by meetings, emails or chats; its methodology note bases that figure on the top 20% of users by ping volume, so it describes the heaviest-loaded users, not the median one. It reports that 57% of meetings are ad hoc calls without a calendar invite (the methodology note gives 60% for the top 20% of users by meeting volume), that 1 in 10 scheduled meetings is booked at the last minute, and that meetings after 8 pm rose 16% year over year. In the same report’s survey of 31,000 workers, 48% of employees and 52% of leaders say their work feels chaotic and fragmented. A plan that cannot absorb a same-day meeting is not a plan for this environment. For the cost side of that synchronous time, see The True Cost of Meetings in 2026.
Third, unpredictability has costs of its own, beyond lost hours. The strongest data come from hourly service work, not knowledge work, so they are a signal rather than a direct measure for desk workers. The Shift Project surveyed 27,792 hourly workers at 80 of the largest US food-service and retail firms between June 2016 and October 2017. About two-thirds received their schedule with less than two weeks’ notice, about one-third with less than one week’s. Workers saw a 32% average gap between their highest-hours and lowest-hours week in the previous month. In the predicted probabilities reported by Schneider and Harknett, 64% of workers who had a shift cancelled reported psychological distress, against 43% of those who did not; 82% against 72% reported poor sleep quality. The design is cross-sectional, so it shows association, not cause. But the direction is consistent: the less predictable the week, the more it costs the person living it.
The conclusion is not “stop planning.” It is “plan differently”: fix the few things that must not move, and stop pretending the rest is fixed.
Where do 168 hours actually go?
The table below converts the 2025 American Time Use Survey (published 25 June 2026) from hours per day into hours per week for employed adults aged 18 and over. ATUS figures are averages over all days of the year, including weekends, holidays and days off, and over all people in the group, including those who did not do the activity that day. Multiplying by seven gives a weekly allocation, not anyone’s actual diary.
CEOtudent analysis of BLS American Time Use Survey 2025, Table 8B (employed, 18 years and over). Weekly figures = published average hours per day x 7.
| Category (ATUS primary activity) | No children under 18: h/day | No children: h/week | Youngest child under 6: h/day | Youngest child under 6: h/week | Weekly difference |
|---|---|---|---|---|---|
| Sleeping (includes naps and sleeplessness) | 8.83 | 61.8 | 8.56 | 59.9 | -1.9 |
| Other personal care (personal care minus sleeping) | 0.75 | 5.3 | 0.73 | 5.1 | -0.1 |
| Eating and drinking | 1.16 | 8.1 | 1.25 | 8.8 | +0.6 |
| Working and work-related activities | 5.44 | 38.1 | 5.30 | 37.1 | -1.0 |
| Household activities | 1.62 | 11.3 | 1.51 | 10.6 | -0.8 |
| Purchasing goods and services | 0.68 | 4.8 | 0.54 | 3.8 | -1.0 |
| Caring for and helping household members | 0.08 | 0.6 | 1.92 | 13.4 | +12.9 |
| Leisure and sports | 4.53 | 31.7 | 3.18 | 22.3 | -9.5 |
| of which: sports, exercise and recreation | 0.32 | 2.2 | 0.29 | 2.0 | -0.2 |
| of which: socializing and communicating | 0.51 | 3.6 | 0.74 | 5.2 | +1.6 |
| All other categories (remainder to 24 hours) | 0.91 | 6.4 | 1.01 | 7.1 | +0.7 |
| Total | 24.00 | 168.0 | 24.00 | 168.0 |
Three things fall out of this table that the colour-coded calendar ignores.
The body takes almost half the week before work starts. Sleeping, other personal care and eating add up to 75.2 hours for employed adults without children, 44.8% of the week. For parents of a child under 6, adding care brings the fixed load to 87.2 hours, 51.9% of the week. After those anchors and work, the no-children group has 54.7 hours left for everything else; the young-child group has 43.7. That second number, not 168, is the real design space.
A young child is mostly paid for out of leisure, not work. Weekly work time differs by only about one hour between the two groups. Care rises by 12.9 hours; leisure falls by 9.5 hours, and most of that fall is television (6.5 fewer hours a week). Any ideal-week template that assumes a fixed pool of “free evenings” will be wrong for a caregiver within days.
Exercise is the thinnest anchor. Sports, exercise and recreation average 0.32 hours a day for the no-children group, which is about 134 minutes a week, and 122 minutes for parents of a child under 6. The WHO 2020 guidelines recommend 150-300 minutes of moderate-intensity (or 75-150 minutes of vigorous-intensity) aerobic activity a week, plus muscle-strengthening on 2 or more days. The two numbers measure different things: the ATUS category includes recreation and counts clock time, while WHO counts intensity and includes activity in every domain the guidelines define, from leisure to occupation, home and transport. So the 16-28 minute shortfall is a prompt, not a measured deficit. The practical reading: movement is the anchor most likely to be cut when a week breaks, so it needs to be the most protected.
One caveat applies to the whole table. BLS suspended all ATUS interviews from 1 October to 12 November 2025 during the federal government shutdown, so there are no diaries for 30 September through 11 November. BLS reweighted the fourth quarter and states that it is not possible to quantify the effect on the 2025 estimates.
Is the picture different in Europe?
The Harmonised European Time Use Surveys (HETUS) 2020 wave, published by Eurostat, allow a check for full-time employed people specifically. Fieldwork years differ by country, so the comparison is between national snapshots taken between 2020 and 2023, not one common year.
Verified data: Eurostat, dataset tus_20selfstat (HETUS 2020 wave), employed persons working full-time, both sexes, time spent in the main activity, mean over all individuals and all days (hh:mm per day). Weekly columns are CEOtudent calculations (x 7).
| Country (fieldwork) | Sleeping | Employment | Travel to/from work | Sports (excl. walking) | Walking and hiking | Leisure, social and associative life | Employment, h/week (calc.) | Sport and walking, min/week (calc.) |
|---|---|---|---|---|---|---|---|---|
| Germany (2022) | 8:21 | 4:37 | 0:25 | 0:18 | 0:13 | 5:02 | 32.3 | 217 |
| Netherlands (2021-2023) | 8:29 | 4:56 | 0:33 | 0:18 | 0:07 | 4:22 | 34.5 | 175 |
| Austria (2021-2022) | 8:18 | 5:20 | 0:32 | 0:23 | 0:15 | 4:50 | 37.3 | 266 |
| Poland (2023) | 8:04 | 5:43 | 0:37 | 0:13 | 0:10 | 3:35 | 40.0 | 161 |
| Finland (2020-2021) | 8:11 | 5:07 | 0:23 | 0:21 | 0:11 | 4:28 | 35.8 | 224 |
| Norway (2022-2023) | 8:00 | 5:21 | 0:24 | 0:28 | 0:11 | 4:46 | 37.4 | 273 |
Source: Eurostat, tus_20selfstat, data updated 29 July 2026; data collection periods from the Eurostat HETUS 2020 metadata. Activity classifications differ from ATUS, so cross-Atlantic comparisons of single rows are not valid.
Two points matter for design. First, even full-time workers average only 32-40 hours of employment per calendar week once holidays, leave and weekends are spread across the year. A plan that books 50 work hours into every week is planning for a year that national diaries do not show. Second, the spread in movement time is wide: from 161 minutes a week of sport and walking in Poland to 273 in Norway, among people with the same full-time status. Country context shapes defaults, which is exactly why defaults should be designed rather than inherited.
What does the evidence say actually helps a plan survive?
Verified data: research findings as printed in each source.
| Question | Finding | Source |
|---|---|---|
| How short do our estimates run? | Best estimate 33.9 days vs actual 55.5 days; 29.7% finished by their best-estimate date; means based on 33 subjects | Buehler, Griffin & Ross (1994), Study 1, Table 1 |
| Do if-then plans help? (original estimate) | d = .65 on goal attainment across 94 independent tests | Gollwitzer & Sheeran (2006), abstract |
| Do if-then plans help? (current estimate) | Sample-weighted d = .36 across 642 tests; trim-and-fill adjusted d = .35; robust Bayesian estimate d = .15 | Sheeran, Listrom & Gollwitzer (2025) |
| Where do they work least? | Behaviour change d = .27; physical activity d = .14 across 64 tests | Sheeran, Listrom & Gollwitzer (2025) |
| What makes them work better? | Larger effects when plans had an if-then format, people were highly motivated, and plans were rehearsed | Sheeran, Listrom & Gollwitzer (2025), abstract |
| What is the sleep floor? | Adults should sleep 7 or more hours per night on a regular basis | AASM and SRS consensus statement (Watson et al., 2015) |
| What is the movement floor? | 150-300 min moderate or 75-150 min vigorous aerobic activity per week, plus muscle-strengthening on 2 or more days | WHO 2020 guidelines (Bull et al., 2020) |
| How common is inactivity? | 31% of the world’s adults, 1.8 billion people, do not meet the 150-minute recommendation | WHO fact sheet, 4 November 2026 |
The implementation-intention row is the one most often misquoted. The 2006 figure of d = .65 is real, but the 2025 meta-analysis with nearly seven times as many tests found a sample-weighted effect a little over half as large, “small according to conventional guidelines” in the authors’ words, with substantial publication bias. If-then planning still works; it is simply not magic, and it works worst for exercise, which is exactly the anchor the time-use data say is thinnest. The design response is to stack several small supports on that anchor (a fixed time, a fixed place, a rehearsed plan, a fallback version) rather than rely on one intention.
How do you build an ideal week that bends without breaking?
The CEOtudent ideal week has three layers.
Layer 1: Anchors. A short list of blocks that do not move when work changes: the sleep window, care obligations, a movement minimum, one or two relationship blocks, and a small number of deep-work blocks. Anchors are defended like a board meeting. They can be re-timed within the day, but not deleted.
Layer 2: Flexible capacity. Collaborative work, reactive work, admin and household tasks. These are planned as hours, not as fixed slots, and they expand or shrink each week. A capacity buffer of unassigned work hours sits inside this layer. The buffer is the shock absorber the planning-fallacy data say you need.
Layer 3: The re-plan ritual. Thirty minutes once a week, using the weekly review linked above, to reset Layer 2 against what next week actually holds. The anchors are checked, not rebuilt.
Below is the fill-in template. The “Reference” column gives the verified figure to sanity-check your number against. The “Example” column is one worked week for a full-time knowledge worker without children; every number in it is an illustrative assumption, not a recommendation, and it sums to 168.
CEOtudent editorial framework: the Ideal Week Template. Example values are illustrative assumptions.
| Block | Purpose | Your hours | Example (illustrative assumption) | Anchor or flexible | Reference figure | AI assist |
|---|---|---|---|---|---|---|
| Sleep window | Recovery; the base every other block depends on | ___ | 56 (8 per night) | Anchor | Floor: 7 h x 7 nights = 49 h (AASM/SRS); ATUS employed, no children: 61.8 h in the sleeping category | None needed; keep devices out of it |
| Personal care and meals | Body maintenance | ___ | 13 | Anchor | ATUS: 5.3 h other personal care + 8.1 h eating | Meal-plan and grocery list drafts |
| Movement | Health minimum | ___ | 4 (240 min) | Anchor | WHO: 150-300 min moderate per week, plus 2 strength days | Builds a fallback 20-minute version of each session |
| Relationship blocks | Named people, named times | ___ | 8 | Anchor | ATUS: 3.6 h socializing (no children) | Reminders for commitments made in conversation |
| Deep-work blocks | The one or two outcomes that move the quarter | ___ | 12 (for example 4 x 3 h) | Anchor (re-timable) | Place them in your peak window; see the chronotype self-test | Prepares source packs and first drafts before each block |
| Collaborative and reactive work | Meetings, messages, requests | ___ | 22 | Flexible | Ad hoc share of meetings: 57% (Microsoft, company-reported) | Summarises threads, drafts replies, batches requests |
| Capacity buffer | Unassigned work hours for the unplanned | ___ | 6 (15% of work hours) | Flexible, held back | Actual time ran 1.64 x best estimate (Buehler et al., 1994) | Compares your planned and actual hours each week |
| Commute and transitions | Moving between places and modes | ___ | 3 | Flexible | HETUS full-time workers: 2.7-4.3 h a week travel to/from work | Route and timing checks |
| Household and admin | Running the household | ___ | 14 | Flexible | ATUS: 11.3 h household + 4.8 h purchasing | Drafts forms, tracks renewals and bills |
| Weekly re-plan | Reset Layer 2 for next week | ___ | 1 | Anchor | Use the weekly review protocol | Pre-fills next week’s calendar conflicts and open loops |
| Unscheduled rest and leisure | Slack that belongs to you | ___ | 29 | Flexible | ATUS: 31.7 h leisure and sports (no children) | None; this is the block AI should not optimise |
| Total | 168 | 168 |
In the example, anchors (sleep, care and meals, movement, relationships and the re-plan) take 82 hours. Work totals 40 hours, of which 6 are buffer. The buffer is set at 15% of work hours as a starting point; that percentage is an editorial choice, not a research finding. The planning-fallacy study measured calendar days to finish a thesis, not hours of effort, so its 1.64 ratio should not be applied mechanically to weekly hours. What it does justify is holding something back, and then setting the size from your own record: after four weeks, compare planned and actual work hours and resize the buffer to match.
Two rules make the template work. Fill the anchors first, in ink; fill the flexible layer last, in pencil. And never book the buffer in advance. A buffer assigned on Sunday is just more plan.
For the placement of deep-work blocks inside the day, use the Chronotypes and Peak Performance self-test and the block design in Ultradian Rhythms Explained. Those tables are not rebuilt here.
What changes for makers, managers, caregivers and freelancers?
The same three layers apply to everyone, but the weight of each layer depends on how work arrives. The variants below are CEOtudent editorial framework, not research findings. They describe where each pattern usually breaks and what to anchor first.
CEOtudent editorial framework: Ideal Week variants by work pattern.
| Work pattern | What arrives unpredictably | Anchor first | Deep-work design | Buffer starting point | Re-plan cadence | Main failure mode |
|---|---|---|---|---|---|---|
| Maker (writer, engineer, analyst, designer) | Reviews, bug reports, requests to “jump on a call” | Two or more half-day deep blocks a week | Long blocks, same days each week, meetings clustered on the other days | 10-15% of work hours | Weekly | Meetings leak into deep blocks one at a time |
| Manager | People problems, escalations, leadership requests | One protected thinking block plus 1:1s as fixed anchors | Shorter blocks (60-90 min) at the start of the day before the queue opens | 20-30% of work hours, because the job is partly reactive by design | Weekly plus a 10-minute daily check | The buffer becomes the job and strategic work disappears |
| Caregiver (child, parent, partner) | Illness, school closures, appointments | Care obligations and sleep; work anchors come second | Fewer, shorter blocks protected by a named back-up person | Keep a buffer on the care side too (back-up cover), not only in work hours | Weekly, with a mid-week checkpoint | Leisure and sleep absorb every shock until they are gone |
| Variable-client freelancer | Client volume, deadlines, payment timing | Sleep, movement and one business-development block | Book deep blocks per project, not per week; re-place them when a project moves | 15-25% of sold hours, to protect delivery quality | Weekly, plus after every new contract | Saying yes to every request in a busy week, then idling in a quiet one |
The caregiver row is where the time-use data matter most. The 12.9 extra weekly hours of care for parents of children under 6 came almost entirely out of leisure. A caregiver’s ideal week therefore starts from fewer free hours and a more volatile load, and the template should be built from the 43.7-hour design space, not from the 54.7 hours available to people without young children.
For freelancers and managers whose weeks swing between heavy and light, the weekly template sits inside a longer cycle. If a sprint has lasted months, the problem is not the week; it is the season, covered in Seasons of Output.
What do you do when the week breaks?
Every week breaks somewhere. The difference between a resilient design and a fragile one is whether the break has a procedure. The protocol below is a CEOtudent editorial framework. Its logic follows the evidence above: anchors protected in order of cost, buffer spent before anchors, and if-then rules written in advance because rehearsed plans work better.
CEOtudent editorial framework: the “When the week breaks” repair protocol.
| Step | What to do | Time | Why |
|---|---|---|---|
| 1. Name the break | Classify it: volume spike, priority change, schedule shock (meeting, travel, illness), or capacity drop (low energy, poor sleep) | 2 minutes | Each type has a different repair; a volume spike is not fixed by an earlier alarm |
| 2. Spend the buffer first | Move the new work into unassigned buffer hours before touching anything else | 5 minutes | The buffer exists for this; spending it is the plan working |
| 3. Re-time, do not delete, the deep block | Move the displaced deep-work block to the next open slot this week, using a pre-written if-then rule | 5 minutes | Plans with an if-then format and rehearsal show larger effects (Sheeran et al., 2025) |
| 4. Protect anchors in a fixed order | If cuts are still needed: leisure first, then household and admin, then collaborative work. Sleep floor, care obligations and a minimum movement session go last | 5 minutes | Sleep and movement have official floors; leisure has the most hours to give |
| 5. Use the minimum versions | Swap a 60-minute workout for its 20-minute version; swap a dinner for a call | 2 minutes | A smaller anchor kept is worth more than a full anchor skipped |
| 6. Renegotiate one commitment | Ask for one deadline, meeting or deliverable to move, with a specific new date | 10 minutes | This is the CEO move: the calendar is a negotiation, not a verdict |
| 7. Log the delta | Write down planned vs actual hours and what broke the week | 3 minutes | This record sizes next week’s buffer; it is the student move |
Write the if-then rules for step 3 before the week starts, not during the break. Three that cover most cases:
- “If a meeting lands on a deep-work block, then the block moves to the next morning slot that is free, not to the evening.”
- “If the week’s buffer is used up by Wednesday, then new requests get a date next week by default.”
- “If sleep falls below the floor two nights running, then the next day starts with the movement minimum and no new commitments.”
The third rule protects a floor. It is not a diagnosis or a treatment for anything; persistent sleep problems are a matter for a clinician.
When the same break type appears three weeks running, it is not a break any more. It is the real shape of the work, and the template should be redrawn at the next re-plan. If the drain is energy rather than hours, the fix is a capacity audit, not a new calendar.
Is your ideal week real? A five-minute scorecard
Score each line 0 (no), 1 (partly) or 2 (yes) at the end of a week. The scorecard is a CEOtudent editorial framework.
| # | Test | Score |
|---|---|---|
| 1 | The sleep window held at or above 7 hours on at least 5 nights | 0 / 1 / 2 |
| 2 | Movement reached the planned minimum, even if only in its short version | 0 / 1 / 2 |
| 3 | At least one relationship block happened at its planned time | 0 / 1 / 2 |
| 4 | At least two-thirds of deep-work hours happened, even if re-timed | 0 / 1 / 2 |
| 5 | The buffer was left unassigned at the start of the week | 0 / 1 / 2 |
| 6 | Every break was handled with the repair protocol, not ad hoc | 0 / 1 / 2 |
| 7 | Planned vs actual work hours were logged | 0 / 1 / 2 |
| 8 | The re-plan happened, in under 45 minutes | 0 / 1 / 2 |
A score of 12-16 means the design is holding. A score of 7-11 means one layer is under-built, usually the buffer. Below 7, the template is describing a week you do not have; redraw it from your logged hours rather than from the ideal.
Where does AI fit, and where should it stay out?
AI helps most in Layer 2 and the re-plan: summarising a week of messages into the decisions that matter, drafting replies, proposing where displaced blocks can go, and comparing planned with actual hours. That turns the student move into a two-minute check.
It helps least, and can do harm, in Layer 1. An assistant that “optimises” the calendar will find the sleep window, the relationship block and the unscheduled rest and treat them as free capacity, because nothing in the data marks them as anything else. The rule is simple: AI may move flexible blocks; only you move anchors. The same telemetry that shows AI-era work spreading into evenings (meetings after 8 pm up 16% year over year, more than 50 messages outside core hours for the average employee, company-reported) is the reason the anchors need a human owner.
For how AI changes what a deep-work block is for, see Deep Work Is Dead.
What this does not mean
- It does not mean the ATUS averages are targets. They describe what large populations do, including people who did nothing in a category that day. Your numbers can and should differ.
- It does not mean if-then plans are useless. The 2025 estimates are smaller than the 2006 one, but the effect remained across 642 tests. Use if-then plans as one support among several, especially for movement.
- It does not apply evenly to shift workers. People whose employer sets their hours with little notice cannot anchor what they do not control. For them, the anchors are the few things they do control (sleep routine around shifts, one fixed relationship block), and the policy debate about advance notice matters more than any template.
- It is not health advice. Sleep and activity figures here are official guideline numbers and research findings, cited as general information, not personal medical guidance.
Frequently asked questions
How many hours a week do employed people actually work?
In the 2025 American Time Use Survey, full-time employed people worked an average of 8.1 hours on the days they worked: 8.5 hours on an average weekday and 5.5 hours on an average weekend day. Spread over all days of the year, employed adults without children average about 38.1 hours a week of work and work-related activity in CEOtudent’s conversion.
What is the difference between time blocking and an ideal week?
Time blocking assigns tasks to slots. An ideal week, in this framework, decides which slots are anchors that never disappear, how many hours stay flexible, and how much capacity is held back. It is a design for a type of week, not a schedule for a specific one.
How big should a capacity buffer be?
Start with 10-30% of work hours depending on how reactive the role is (see the variants table), then resize it after four weeks using your own planned-versus-actual record. Those starting ranges are editorial; the research only shows that best-guess plans run short.
Do implementation intentions really work?
Yes, but modestly. The 2025 meta-analysis of 642 tests found a sample-weighted effect of d = .36, smaller after correction for publication bias, and d = .14 for physical activity. Effects were larger with an if-then format, high motivation and rehearsal.
How much sleep and exercise should the anchors protect?
The AASM and Sleep Research Society consensus recommends 7 or more hours per night for adults on a regular basis. The WHO 2020 guidelines recommend 150-300 minutes of moderate-intensity aerobic activity a week, or 75-150 minutes vigorous, plus muscle-strengthening on 2 or more days. This is general information, not medical advice.
What if an employer controls the schedule?
Anchor what you control and treat the rest as flexible capacity. The Shift Project data show that about two-thirds of the hourly service workers surveyed got less than two weeks’ notice of their schedules, and that unpredictable schedules were associated with more distress and poorer sleep.
How often should the template itself be redesigned?
Re-plan the flexible layer weekly. Redraw the template when the same break type appears three weeks in a row, when your work pattern changes (new role, new client mix, a new child), or at your annual review.
Sources
- U.S. Bureau of Labor Statistics. American Time Use Survey Summary, 2025 Results. News release USDL-26-1022, 25 June 2026 (including the notice on the October-November 2025 federal government shutdown).
- U.S. Bureau of Labor Statistics. American Time Use Survey 2025, Table 4: Employed people working and time spent working on days worked, 2025 annual averages.
- U.S. Bureau of Labor Statistics. American Time Use Survey 2025, Table 8B: Time spent in primary activities for the civilian population 18 years and over by presence and age of youngest household child and sex, employed, 2025 annual averages.
- Eurostat. Time spent in the main activity by sex and self-declared labour status (tus_20selfstat), Harmonised European Time Use Surveys 2020 wave. Data updated 29 July 2026.
- Eurostat. Time use survey 2020 (tus_20), reference metadata (ESMS), sections 3.4 and 3.8.
- Buehler, R., Griffin, D., and Ross, M. (1994). Exploring the “planning fallacy”: Why people underestimate their task completion times. Journal of Personality and Social Psychology, 67(3), 366-381.
- Gollwitzer, P. M., and Sheeran, P. (2006). Implementation intentions and goal achievement: A meta-analysis of effects and processes. Advances in Experimental Social Psychology, 38, 69-119 (abstract, University of Konstanz repository record).
- Sheeran, P., Listrom, O., and Gollwitzer, P. M. (2025). The when and how of planning: Meta-analysis of the scope and components of implementation intentions in 642 tests. European Review of Social Psychology, 36(1), 162-194.
- Watson, N. F., et al., Consensus Conference Panel (2015). Recommended amount of sleep for a healthy adult: A joint consensus statement of the American Academy of Sleep Medicine and Sleep Research Society. SLEEP, 38(6), 843-844.
- Bull, F. C., et al. (2020). World Health Organization 2020 guidelines on physical activity and sedentary behaviour. British Journal of Sports Medicine, 54(24), 1451-1462.
- World Health Organization. Physical activity fact sheet, 4 November 2026.
- Schneider, D., and Harknett, K. (2019). Consequences of routine work schedule instability for worker health and wellbeing. The Shift Project research brief, February 2019 (summarising the authors’ paper in American Sociological Review).
- Microsoft WorkLab. Breaking down the infinite workday. Work Trend Index Special Report, 17 June 2025 (company-reported; Microsoft 365 telemetry and a survey of 31,000 workers in 31 markets).
Note on method: The 168-hour budget table, the weekly columns of the HETUS table, the planning-fallacy ratios, the Ideal Week Template, the variants table, the repair protocol and the scorecard are CEOtudent analyses or CEOtudent editorial frameworks; every derived number is recomputed in a script kept with this article. The example week uses illustrative assumptions. We deliberately did not use the 2006 implementation-intention effect (d = .65) as a current estimate, because the 2025 meta-analysis of 642 tests replaces it. We did not use the widely repeated “one in three adults is short of sleep” figure, because we could not open its primary source for this piece. Microsoft figures are company-reported.
This content was compiled with the support of AI following in-depth research, then written and prepared for publication by the CEOtudent editorial team.
This post is also available in:
