A time audit records where working time actually goes for one or two weeks, then compares it with where you thought it went. To run one for a team: pick the question, record every block of time by category, total it, compare it with your priorities and change one or two things. Automatic tracking makes the recording cheap. The value is in the comparison.
Key takeaways
- Start with a question, such as "why are projects running late?"
- Two weeks is enough to see the pattern.
- Six to eight categories. More than that and the log gets filled in badly.
- End with one or two changes, then audit again in a quarter.
When is a time audit worth doing?
Work is late
The team is busy and delivery still slips.
Capacity is unclear
You cannot say whether you can take on another client.
Pricing feels wrong
Fixed-fee work keeps costing more than quoted.
People are stretched
Long hours, and nobody can say on what.
How do you run a time audit?
Pick one question
"How much of the week goes to client work?" is a question. "Let's see what people do" is not.
Choose the categories
Six to eight, the same for everyone. Use the table below as a start.
Record for two weeks
Every working block gets a category. Use automatic tracking if you have it.
Total by category
Hours and share of the week, for the team and for each role.
Compare with your priorities
Put the actual split next to the split you would choose.
Change one or two things
Then run the audit again in three months to see whether it worked.
Which categories should you use?
| Category | What goes in it |
|---|---|
| Core work | The work the role exists to do: design, code, cases, calls with customers |
| Meetings | Scheduled calls and meetings, internal and external |
| Communication | Email and chat |
| Admin | Timesheets, expenses, approvals, reporting |
| Support and interruptions | Unplanned requests and fixes |
| Planning and learning | Estimating, training, reading |
| Breaks | Short breaks during the day |
Keep the list short. The aim is a picture you can act on, not a perfect ledger.
What does a result look like?
Share of time = hours in a category ÷ total hours × 100
Example: 9 hours of meetings in a 40-hour week = 22.5%
- Core work18h, 45%
- Meetings9h, 22.5%
- Email and chat7h, 17.5%
- Admin4h, 10%
- Other2h, 5%
An illustration, not a benchmark. Your own numbers are the only ones that matter.
Most managers guess the first bar too high. Seeing less than half the week on core work is common, and it is the finding that makes the audit worth doing.
What questions should you ask of the data?
Questions to work through
- What share of the week goes to the work we are paid for?
- Which category is larger than anyone expected?
- Which meetings could be shorter, rarer or written updates?
- Who carries the unplanned work, and is that fair?
- Which hours of the day are least interrupted?
- What would we stop doing if we had to free five hours a week?
If communication and meetings dominate, read context switching at work.
Manual log or automatic tracking?
| Manual log | Automatic tracking | |
|---|---|---|
| Effort | Every block written down by hand | Recorded in the background |
| Accuracy | Depends on discipline and memory | Records what happened |
| Blind spots | Small tasks get left out | Work away from the computer, added by hand |
| Length you can sustain | One or two weeks | Continuous |
| Best for | A one-off audit with no tooling | Repeating the audit every quarter |
A manual time log to copy
Date,Start,End,Minutes,Category,Project or client,What I did,Planned or unplanned 2027-03-08,09:00,09:25,25,Communication,,Email and chat catch-up,Planned 2027-03-08,09:25,11:05,100,Core work,Website redesign,Homepage layout,Planned 2027-03-08,11:05,11:35,30,Meetings,Website redesign,Client check-in,Planned 2027-03-08,11:35,11:50,15,Support and interruptions,Support retainer,Urgent fix request,Unplanned
Total the minutes by category at the end of each week.
What mistakes spoil a time audit?
Do
- Tell the team the question and what you will do with the answer
- Audit roles and teams, not individuals
- Share the results with everyone who took part
- Pick one or two changes and follow through
Do not
- Run it without saying why
- Use it to find the "least productive" person
- Collect data for months and change nothing
- Treat thinking and breaks as waste
How does BirdsEye make an audit easier?
BirdsEye records the working day automatically, so the audit becomes a matter of reading reports.
- Application usage. How much time went to each application and website domain.
- Productivity split. Productive, neutral and unproductive time, by the rules you set.
- Time by project. Where project hours went, against budget.
- Hours by day of week and hour-of-day heat map. When the work happens.
- Period comparison. Each figure against the previous period, so the second audit is one click.
- Saved views. Keep the filter you used and recall it next quarter.
Analytics with period comparison and saved views are on the Team plan and above. Employees can see their own data too. See reports and analytics.
Frequently asked questions
What is a time audit?
A short exercise that records how working time is actually spent and compares it with how you intended to spend it.
How long should a time audit last?
Two weeks is enough for most teams. One week can be distorted by a deadline or a holiday.
How often should you run one?
Once a quarter is plenty. Run another after any large change in workload, team or tools.
Should employees know about a time audit?
Yes. Tell them the question, the categories and what will happen with the results. An audit people do not understand produces data you cannot trust.