Summary
A time audit is a two-week exercise that answers one question: where does the team's working time actually go? Not where people think it goes, which is reliably wrong, and not where the job description says it should. The answer is usually surprising, and the surprises are where the recoverable hours are. This guide gives the five-step method, the category scheme that makes results comparable, the three patterns that show up in almost every audit, and the fixes that follow from each.
E-Monitor runs the collection step automatically: time per application, site and project, categorised, for the whole team, with nothing to fill in. Sign up →
Why Self-Reported Time Is Wrong
Ask a team to estimate their week and the estimates will add to about sixty hours each, with meetings undercounted, email undercounted, and "real work" overcounted. This is not dishonesty; memory compresses interruptions and expands the work people find meaningful. A time audit exists because the self-report cannot be trusted, and the gap between estimate and measurement is itself the first finding.
The method below works with either manual logs or automatic collection. Manual logs are free and painful, and their accuracy decays after about three days. Automatic collection through time tracking and app categorisation removes the effort and the decay, which is why most teams that run a second audit switch to it.
Step 1: Define the Categories Before You Start
The audit is only as useful as its categories, and they must be fixed before collection begins so that everyone sorts time the same way. Six top-level categories cover most knowledge work:
- Core work: the output the role exists to produce
- Coordination: meetings, stand-ups, planning
- Communication: email, chat, messages, outside meetings
- Administration: timesheets, expenses, reporting, tool admin
- Learning: training, documentation, research
- Unallocated: idle, switching, waiting
Add one or two role-specific subcategories under core work, such as "client delivery" and "internal projects", but resist more than that. Ten categories produce arguments about sorting; six produce findings.
Step 2: Collect for Two Full Weeks
One week is not enough because weeks vary; two consecutive weeks smooth out a single bad Monday. Collect across the whole team at once, so that the results are comparable and nobody feels singled out. Tell the team what the audit is for and that the results will be shared with them, category by category, before any decision is made.
If you are collecting manually, ask for entries at least every two hours and accept that the second week will be less accurate. If you are collecting automatically, check the category mapping after day two and correct the applications that landed in the wrong bucket; misclassified tools are the most common source of a misleading audit.
Step 3: Build the Three Views
Once the two weeks are in, produce three tables and nothing else.
Team split by category. Percentage of total time in each of the six categories. This is the headline. In most knowledge teams core work lands between forty and fifty-five percent, and anything below forty is the first thing to explain.
Split by person, anonymised for the team view. The range within a category matters more than the average. If coordination time runs from ten percent for one person to forty for another in the same role, the job is being done two different ways.
Time by day and hour. Where in the week the core work happens, and where the fragmentation is. This view usually reveals the meeting wall that leaves no two-hour block on certain days.
| Category | Typical range | Investigate if |
|---|---|---|
| Core work | 40 to 55% | Below 40% |
| Coordination | 15 to 25% | Above 30% |
| Communication | 10 to 20% | Above 25% |
| Administration | 5 to 10% | Above 12% |
| Learning | 3 to 8% | Below 2% |
| Unallocated | 5 to 12% | Above 15% |
Step 4: Find the Three Usual Patterns
The meeting wall
Coordination above thirty percent, with the day-and-hour view showing no uninterrupted block longer than an hour on most days. The fix is structural: protected focus blocks, a meeting-free morning, and a review of recurring meetings against the meeting overload guide. Teams typically recover three to five hours a week each from this alone.
The hidden admin job
Administration above twelve percent, concentrated in one or two people. Someone is assembling a report by hand, reconciling two systems, or answering the same question for everyone. The fix is usually a template, an automation or a shared document, and it frees a day a fortnight for the person carrying it.
The fragmentation tax
Core work near the healthy range but delivered in slices under twenty minutes, visible in the hour view as constant switching. The context switching guide covers the cost; the fix is batching communication into two or three windows a day and turning off notifications between them.
Step 5: Share, Decide, Re-Audit
Present the three views to the team before deciding anything. Ask what surprised them and what they would change; the person carrying the hidden admin job usually already knows the fix. Pick two changes, not ten, and write down what you expect the category split to look like in three months.
Then re-audit at the three-month mark with the same categories. The second audit is where the value is: it shows whether the changes stuck and gives you a before-and-after in hours that you can put in front of anyone. The utilization rate guide shows how to turn the core-work percentage into a number leadership already tracks.
A Worked Example: A Twelve-Person Support Team
A software company ran the method above on its customer support team, which believed it was spending most of its week on tickets. The team lead's estimate was seventy percent core work, fifteen percent meetings, the rest scattered. Two weeks of automatic collection said otherwise.
| Category | Estimated | Measured | Gap |
|---|---|---|---|
| Core work (tickets) | 70% | 46% | -24 |
| Coordination | 15% | 27% | +12 |
| Communication | 5% | 14% | +9 |
| Administration | 5% | 9% | +4 |
| Learning | 3% | 1% | -2 |
| Unallocated | 2% | 3% | +1 |
The person view explained the coordination number: three senior agents were in twenty-seven hours of meetings a fortnight each, most of them escalation reviews that the rest of the team never attended. The hour view showed that Tuesday and Thursday mornings had no ticket work at all for those three, because the reviews were booked back to back. The administration figure came from one agent who spent four hours a week producing a weekly quality report by copying ticket data into a spreadsheet.
The team chose two changes: escalation reviews moved to a single ninety-minute slot on Wednesday with a written agenda, and the quality report was replaced with a saved dashboard view. Nothing else changed. At the three-month re-audit, core work had risen to fifty-eight percent, coordination had fallen to sixteen, and the report author had a day a week back. Learning time doubled without anyone planning it, because the freed hours went somewhere. The whole exercise cost the team lead about six hours across the two audits.
Time Audit Template
The audit fits on one page. Copy the checklist, fill in the dates, and run it in order.
- Purpose: one sentence on what the team wants to learn
- Categories: the six above, plus at most two subcategories
- Collection method: manual log or automatic tracking, decided before day one
- Window: two consecutive full weeks, dates fixed
- Announcement: what is measured, what is not, who sees it, sent a week ahead
- Day-two check: category mapping reviewed and corrected
- Three views: team split, anonymised person range, day-and-hour heatmap
- Team review: findings shared before any decision
- Two changes: chosen with the team, expected effect written down
- Re-audit date: three months out, same categories
Keeping the Audit Honest
A time audit is a diagnostic of the work, and it stops working the moment people believe it is a measurement of them. Keep the person-level view anonymised in anything shared, share every finding with the team before acting on it, and never use audit data in a performance conversation. If the team trusts the audit, they will tell you where the hours go; if they do not, they will spend the two weeks performing, and you will audit the performance.
Frequently Asked Questions
1. What is a time audit?
A two-week exercise that measures where a team's working time actually goes, sorted into a fixed set of categories such as core work, coordination, communication and administration, so the gap between assumed and real time use becomes visible and fixable.
2. How long should a time audit run?
Two consecutive full weeks. One week is distorted by whatever happened that week; two smooth it out. Re-audit with the same categories three months after making changes.
3. Should I use manual logs or automatic tracking?
Manual logs work for a first audit but decay in accuracy after about three days. Automatic time and application tracking removes the effort and the decay and gives comparable data across the whole team; most teams switch to it for the second audit.
4. What percentage of time should be core work?
In most knowledge-work teams, forty to fifty-five percent. Below forty is the first thing to investigate, and it is usually explained by meeting load or a hidden administrative task rather than by individual effort.
5. How much time does a time audit usually recover?
Teams typically recover four to six hours a week per person from the first two changes, most of it from cutting recurring meetings and automating one manual administrative process.
6. Can time audit data be used for performance reviews?
It should not be. The audit works only when people trust it as a diagnostic of the work. Using it to judge individuals turns the next audit into a performance, and the data stops being true.
Run the audit without the spreadsheets E-Monitor categorises time by application, site and project for the whole team automatically, so the collection step takes no effort and the re-audit is one click. Sign up →
