The most instructive time audit is a specific one. Principles are useful; watching someone work through the process with real (or representative) data shows you what the experience actually looks like.
This case study follows Maya, an engineering manager at a mid-sized software company, through a complete 7-Day Time Audit using AI. Her situation is common: she’s effective at her job, has a full calendar, and feels the constant press of too much to do. But she doesn’t know, with data, where her time is actually going.
The Setup
Maya manages a team of eight engineers. Her role includes regular one-on-ones, sprint planning, stakeholder meetings, and the ongoing work of unblocking her team. She also has a personal goal of spending more time on technical strategy — a part of her role she believes is neglected.
Before the audit, she estimated her week as:
- Meetings: 15-18 hours
- Email and Slack: 3-4 hours
- Deep strategy work: 4-5 hours
- One-on-ones and coaching: 3-4 hours
- Miscellaneous admin: 2 hours
She considered herself reasonably aware of how she spent her time. She’d thought about it enough to have a rough picture.
The Logging Week
Maya used a spreadsheet template with 30-minute slots pre-filled for the full waking day (6:30am to 10:30pm, 32 slots per day). She logged in real time, noting activity, context, and energy on a 1-3 scale.
She found the logging mildly annoying for the first two days, then habitual by the third. The biggest friction point was remembering to log during heavy-meeting stretches — she missed several afternoon blocks on Tuesday and had to reconstruct them, which she noted in the log.
The log ran Monday through Sunday. She included weekends because she wanted the full picture, not just the workweek.
The AI Categorization
At the end of seven days, Maya had 212 logged entries. She pasted the raw log into Claude with the following prompt:
I'm an engineering manager at a software company. I've completed a 7-day time audit.
My primary question: am I spending time in proportion to what my role actually requires?
Please categorize these entries into a structure appropriate for a manager role
and produce a summary table with hours and percentage of waking time.
[raw log pasted]The AI proposed a category structure and returned a categorized summary. Maya reviewed the ambiguous entries — there were fourteen — and confirmed or corrected each one.
The Findings
The actual allocation, per the audit:
| Category | Hours | % Waking Time |
|---|---|---|
| Meetings (group) | 22.5 | 20% |
| One-on-ones and direct coaching | 6.0 | 5% |
| Email and Slack | 11.5 | 10% |
| Deep strategy/technical work | 2.0 | 1.8% |
| Reactive unblocking (urgent ad-hoc) | 5.5 | 4.9% |
| Planning and review | 3.0 | 2.7% |
| Personal obligations | 14.0 | 12.5% |
| Sleep | 47.5 | 42% |
| Exercise | 3.0 | 2.7% |
| Recovery and leisure | 10.5 | 9.4% |
| Transition and unclear | 8.5 | 7.6% |
| Total waking hours | 112 |
The findings that surprised her:
Meetings ran 22.5 hours, not 15-18. The gap between her estimate and the actual data was approximately five to seven hours. When she looked at the breakdown, the difference came from meetings that ran long (an average of fifteen minutes over scheduled time), recurring syncs she’d mentally filed as “fifteen minutes” that consistently ran thirty, and two ad-hoc calls she hadn’t counted at all.
Email and Slack consumed 11.5 hours — more than three times her estimate. This was the largest shock. She had estimated three to four hours. The actual pattern: she checked messages every fifteen to twenty minutes throughout the day, with each check consuming five to fifteen minutes. The cumulative total was more than two working days per week.
Deep strategy work was 2 hours total. Against her estimate of four to five hours, and against her stated priority of doing more of it.
Transition time was 8.5 hours. She hadn’t tracked this consciously — it emerged from the audit as the accumulated time between activities where no single activity dominated.
The Gap Analysis
Maya ran the gap analysis prompt:
Here is my categorized summary. My current priorities:
1. Technical strategy and architecture decisions (high leverage, currently neglected)
2. Direct coaching of senior engineers on the team
3. Reducing meeting overhead on my team
Given this, please:
1. Identify the 2-3 most significant gaps
2. Distinguish lost time from necessary recovery
3. Give me hypotheses about why the gaps exist
4. Suggest what one change would have the highest impactThe AI’s analysis identified three gaps:
Gap 1: Deep strategy work is nearly absent. At 2 hours out of 112 waking hours, and with it listed as her highest priority, this is the most significant misalignment. The AI’s hypothesis: strategy work requires uninterrupted blocks of 60-90 minutes minimum. Given the fragmentation of her calendar (most calendar blocks are 30 minutes or less, and many are back-to-back), there’s no structural space for this work to happen.
Gap 2: Email and Slack are consuming time on a fragmented, reactive basis. The 11.5 hours is not the core problem — knowledge work at her level does require significant communication time. The problem is the distribution: checking every fifteen to twenty minutes means she never gets the psychological distance from communication needed to focus on anything else. The AI distinguished this from her recovery time, which it assessed as adequate, and from her one-on-ones, which it flagged as potentially underinvested relative to her coaching priority.
Gap 3: Transition time is substantial and structural. 8.5 hours of unclear, between-activity time is almost certainly an artifact of context switching. Each meeting, urgent message, or reactive task triggers a transition that takes time to recover from — and those transitions accumulate.
On the question of lost time versus necessary recovery: the AI noted that her leisure and sleep time appeared to be adequate by evidence-based standards (Sonnentag’s recovery framework suggests that sustained detachment from work during off-hours predicts next-day performance, and Maya’s evening leisure time appeared genuinely low-effort and restorative). It specifically flagged that cutting sleep or leisure to gain work hours would likely produce a net performance loss.
The Three Changes
Maya committed to three specific changes after the audit. She chose three rather than one because they were structurally related — each created conditions for the others to work.
Change 1: Two protected strategy blocks per week. Every Tuesday and Thursday, 8:00-10:00am: no meetings booked, phone on do not disturb, Slack notifications snoozed. These two-hour blocks represent her most reliable high-energy time (from the energy ratings in her audit, mornings consistently rated 2-3, afternoons 1-2).
This required negotiating with her calendar — several recurring syncs had been placed in this time. She moved them or made them async.
Change 2: Message batching at three designated times. Email and Slack checked at 8:00am (before the strategy block), 12:00pm, and 5:00pm. Outside those windows, notifications off on all devices. This was the hardest change to implement — it required communicating the expectation to her team and her manager, and setting up auto-responses for urgent queries.
Expected reduction: from 11.5 hours to 5-6 hours. This is not primarily a time saving — it’s a fragmentation reduction. Fewer switches means more sustained focus during work blocks.
Change 3: Meeting audit. She committed to reviewing every recurring meeting and either: cutting it, shortening it, or converting it to async. She ran this as a separate AI-assisted exercise. Net reduction: five hours of recurring meetings per week.
Using Beyond Time for the Redesigned Schedule
After the gap analysis, Maya used Beyond Time (beyondtime.ai) to translate these three changes into a concrete new weekly plan. The AI in Beyond Time took her category summary, her stated priorities, her energy patterns, and the three changes and produced a redesigned weekly schedule that blocked deep work in the identified high-energy windows, clustered meetings in the lower-energy afternoon slots, and built transition time into the calendar explicitly rather than letting it happen by default.
The Follow-Up
Six weeks after implementing the changes, Maya ran a lighter three-day targeted audit.
Deep strategy work had increased from 2 hours to 7.5 hours for the three-day period, suggesting roughly 17-18 hours per week of actual strategic work time — still below what she’d originally estimated before the first audit, but a significant shift.
Email and Slack time had dropped to approximately 5 hours in three days. The batching habit had taken two weeks to feel natural and required one difficult conversation with a direct report who felt cut off by her slower response time.
Meeting time was down to 14 hours projected for the week — still substantial, but reduced.
The finding she reported as most significant: the transition time had dropped from 8.5 hours to approximately 3 hours over the same seven-day period. She hadn’t explicitly targeted transition time as a goal. It improved as a side effect of having larger, more intentionally structured blocks. When the blocks are designed, the transitions shrink.
Your action: Before running your own audit, write down your estimates for each major category: how many hours per week do you think you spend on deep work, email, meetings, and recovery? Be specific. Then run the audit and compare. The gap between your estimates and the actual data is the most useful piece of information the audit produces.
Tags: time audit case study, manager time management, AI time audit, deep work, schedule redesign
Frequently Asked Questions
-
Is this case study based on a real person?
This case study is a composite based on patterns commonly observed in time audit analyses for knowledge workers in manager roles. The specific numbers, findings, and changes are representative of typical audit outcomes for this profile — not a single individual's data. The prompts and process are exactly as described.
-
Are these findings typical for managers?
The broad patterns — high meeting load, fragmented deep work, underestimated email time — are consistent with research on knowledge worker time use and are frequently surfaced in manager time audits. The specific percentages vary by organization, role, and individual, but the direction of the findings is common.