How to Calculate Idle Time: Formulas, Examples and What They Miss

Idle time is scheduled time minus productive time. The formula is Idle Hours = Scheduled Hours − Productive Hours, and idle time percentage is Idle Hours ÷ Scheduled Hours × 100. The number your time tracking software reports as idle is calculated differently — it fires after a set period without keyboard or mouse input — and the two figures are not interchangeable.

Five idle time metrics, with formulas

A check on your arithmetic: idle time percentage and utilization rate always sum to 100%. If they do not, one of the inputs is wrong.

Two neighbouring metrics answer related questions: what is schedule adherence measures whether people worked the hours they were scheduled for, and an attendance sheet records whether they were there at all. Idle time sits between them — present, scheduled, and not producing.

A worked example

A team of 5 people is scheduled for 40 hours each in a week. Time records show 168 productive hours across the team. Average loaded labour cost is $32 an hour.

Scheduled hours    = 5 × 40              = 200
Idle hours         = 200 − 168           = 32
Idle time %        = 32 ÷ 200 × 100      = 16%
Utilization rate   = 168 ÷ 200 × 100     = 84%
Idle per employee  = 32 ÷ 5               = 6.4 hours
Idle labour cost   = 32 × $32             = $1,024

Check: 16% + 84% = 100%. ✓

One caution before acting on that $1,024. It is only a cost if the idle hours were avoidable. Paid breaks, training, setup and unavoidable waiting all land in the same bucket unless you separate them.

Separate them and the number usually collapses. If 22 of those 32 hours were scheduled breaks, training and handover, the avoidable remainder is 10 hours — $320, not $1,024. That is the figure worth taking to a planning meeting, and it is roughly a third of the one the formula hands you.

How to calculate idle time in Excel

Basic: if total time is in A2 and active time in B2, idle time is =A2-B2.

If the cells are formatted as time, multiply by 24 to get decimal hours:

=(A2-B2)*24

The five metrics in a row, with scheduled hours in B2, productive hours in C2 and hourly cost in D2:

A common Excel error here: a total formatted as time rolls over at 24 hours, so 200 hours displays as 8:00. Format the cell as [h]:mm with the square brackets, or convert to a number.

If you are building this from scratch rather than from existing records, a weekly timesheet template gives you the scheduled-hours column already laid out, and best apps to track work hours covers where the productive-hours column comes from.

Idle time versus downtime

Both mean the person or the asset is not producing. They are tracked separately because they have different causes and different answers.

The practical difference: idle time is a planning problem, downtime is a failure. Counting them together hides both.

One line on machine idle time, then we will leave it alone: a machine idle rate uses the same arithmetic against a different denominator, it belongs to maintenance rather than to scheduling, and mixing it with a labour idle rate produces a number that means nothing to either owner. Everything below is about people.

Planned and unplanned idle time

Most teams skip this split, and it is the reason idle time numbers get argued about. A 16% idle rate that is mostly paid breaks and training is a normal week. The same 16% that is mostly waiting on approvals is a process problem. The arithmetic cannot tell you which one you have; only the split can.

Idle time in tracking software is a different number

This is where most idle time conversations go wrong.

The formulas above compare two measured quantities: hours scheduled against hours productive. Time tracking software does something else entirely.

How idle detection actually works

A tracker watches for keyboard and mouse input. When there has been none for a set period — the idle threshold — it marks the time from that point as idle, and usually prompts the user when they return.

That is not a subtraction. It is a threshold firing. And it measures one thing only: whether the input devices were touched.

Why the threshold changes the answer

The same week produces different idle numbers depending on a setting.

A team reporting 20% idle at a three-minute threshold and 6% at fifteen has not changed how it works. The tool changed — the same day, read 3.3× differently.

Before comparing idle numbers between teams, tools or time periods, check that the threshold is the same. It rarely is.

What detected idle time is not

The single most common mistake with this metric: treating detected idle as unproductive time.

Detected idle includes all of the following, and none of it is waste:

• Reading a document, a specification or a long email thread
• Thinking through a problem before typing
• A phone call taken away from the desk
• A meeting attended without a laptop
• Reviewing something printed
• A conversation with a colleague
• Watching a build, a render or a long-running process

What detected idle actually tells you: the input devices were not touched. That is the whole of it. Everything beyond that is inference, and the inference is wrong often enough to matter.

How to use it properly:

1. Read it as a trend, not a judgement. One person’s Tuesday means nothing; a department whose idle rate doubled over a quarter is worth asking about.
2. Never use it alone for a performance decision. Pair it with output.
3. Let people see and annotate their own idle time. A tracker that lets someone mark two hours as “client call” produces better data than one that guesses.
4. Expect roles to differ. A developer and a data entry clerk will never have comparable idle rates, and that difference says nothing about either.

Monitask shows people their own activity data, and lets them annotate it, because a number the person can correct is worth more than one they cannot see.
See how Monitask reports activity

Is idle time paid?

The question every guide skips, and the one with money attached. Detected idle time and unpaid time are not the same thing, and treating them as the same creates real exposure.

Under the Fair Labor Standards Act, whether waiting time counts as hours worked turns on whether the employee was engaged to wait or waiting to be engaged (29 CFR 785.14). The phrase is the regulation’s own, taken from a 1944 Supreme Court decision.

What decides it is not where the person was sitting. This is worth saying plainly, because the usual shorthand — that someone has to be kept on the premises for waiting time to count — is contradicted by the regulation. 29 CFR 785.15 provides that waiting time is worktime “even though the employee is allowed to leave the premises or the job site during such periods of inactivity”. What makes it worktime is that the employee “is unable to use the time effectively for his own purposes”, and that the time “belongs to and is controlled by the employer”.

The regulation’s own examples carry it better than a paraphrase would. A stenographer reading a book while waiting for dictation, a messenger doing a crossword while awaiting assignments, a factory worker talking to colleagues while waiting for machinery to be repaired: all of them, says 29 CFR 785.15, “are working during their periods of inactivity”. Waiting is an integral part of those jobs.

The mirror image is 29 CFR 785.16, which treats time as off duty only where the employee is “completely relieved from duty” for a period “long enough to enable him to use the time effectively for his own purposes” — and adds that they must be “definitely told in advance” that they are free to go.

Short breaks are not a grey area. 29 CFR 785.18 describes rest periods “running from 5 minutes to about 20 minutes” and is unambiguous about what follows: “They must be counted as hours worked.” The Department of Labor applies that as a bright line, stating in its telework guidance that “Breaks of twenty minutes or less must be counted as hours worked.” And the same regulation closes the obvious workaround: “Compensable time of rest periods may not be offset against other working time such as compensable waiting time or on-call time.”

Meal periods work differently, and the difference is the instructive part. A meal break is unpaid only where the employee is completely relieved from duty, and 29 CFR 785.19 is explicit that an employee “is not relieved if he is required to perform any duties, whether active or inactive, while eating”. Duration is evidence, not the test — the regulation says only that “ordinarily 30 minutes or more” is long enough.

None of this changes when the work is remote. The Department’s telework bulletin states that these principles “apply regardless of whether the work is performed at the employer’s worksite, at the employee’s home, or at some other location”.

Why an idle figure cannot simply be deducted

Three reasons, and they compound.

Inactivity is not the test. Every employee in 29 CFR 785.15’s examples is inactive by any tracker’s definition — reading, chatting, doing a crossword — and the regulation says they are working.

An employer may not quietly discount small periods. 29 CFR 785.47 provides that an employer “may not arbitrarily fail to count as hours worked any part, however small, of the employee’s fixed or regular working time”. A threshold applied automatically, without looking at the particular case, is close to the definition of arbitrary.

The record that matters is yours, not the tool’s. 29 CFR 516.2(a)(7) requires the employer to keep records of hours worked each workday and each workweek. A vendor’s idle metric is not that record, and pointing at a dashboard is not the same as keeping one.

There is exactly one narrow route from a break to a pay consequence, and it is not a regulation. It is the Department’s stated position that an unauthorised extension of an authorised break need not be counted — but only where the employer has “expressly and unambiguously communicated” beforehand that the break has a fixed length, that exceeding it is against the rules, and that exceeding it will be punished. All three, communicated in advance. That is a policy you wrote last month, not an inference you drew from a dashboard this morning, and it reaches only the extension: the authorised part of the break stays paid.

There is also a turn here that is easy to miss. Running a tracker makes your position on this worse, not better. The regulations count work the employer “knows or has reason to believe” is being performed (29 CFR 785.11), management “cannot sit back and accept the benefits without compensating for them” (29 CFR 785.13), and the Department’s telework guidance asks what an employer “should have known”. A tool that records activity is evidence that you knew. That is a good reason to run one openly and a bad reason to run one quietly.

If rounding is what brought you here, the 7-minute time clock rule covers that separately.

This is general information, not legal advice. Consult counsel for your circumstances.

Common measurement mistakes

1. Counting paid breaks as waste. Breaks are scheduled, paid and part of the plan. Exclude them before calculating avoidable idle.

2. Using presence as productive time. Productive hours means hours of useful output, not hours logged in. A fully present team can have substantial idle time.

3. Comparing across different thresholds. Covered above. The most common cause of a meaningless comparison.

4. Comparing across different roles. Idle rate is only comparable within similar work.

5. Mixing machine and labour idle time. Different owners, different answers. A machine idle rate belongs to maintenance; a labour idle rate belongs to scheduling.

6. Acting on a single period. One week is noise. Three months is a signal.

How to reduce idle time

Before reducing anything, split the number. Planned idle stays. Unplanned idle divides into causes you control — workload distribution, approval delays, unclear briefs, missing access — and causes you do not. Work only on the first group.

In the worked example above, that is the difference between a $1,024 problem and a $320 one. The second is solvable. The first was never a problem.

The causes worth checking first, in the order they usually turn out to matter:

1. Approval and handover delays. The most common addressable cause, and the one that shows as idle time for the person waiting rather than for the person holding things up.
2. Uneven workload distribution. Idle time concentrated in one or two people is a scheduling problem, not a performance one.
3. Unclear briefs. Someone who cannot start because they do not know what finished looks like registers as idle.
4. Missing access. Credentials, licences, permissions. Cheap to fix and invisible until someone measures it.
5. Genuine troughs in demand. Not addressable by management attention, and worth naming as such so nobody spends a quarter trying.

Sources

• 29 CFR Part 785, subparts B and C — hours worked, waiting time, rest and meal periods. In particular 29 CFR 785.14 (engaged to wait), 29 CFR 785.15 (on duty), 29 CFR 785.16 (off duty), 29 CFR 785.18 (rest) and 29 CFR 785.19 (meal)
• 29 CFR 785.47 — the limit on discounting small periods of working time
• 29 CFR 516.2(a)(7) — the employer’s obligation to record hours worked each workday and workweek
• US Department of Labor, Field Assistance Bulletin No. 2023-1, Telework Under the Fair Labor Standards Act and Family and Medical Leave Act (9 February 2023)
• US Department of Labor, Field Assistance Bulletin No. 2020-5 (24 August 2020) — reasonable diligence in tracking teleworking hours
• US Department of Labor, Wage and Hour Division, Breaks and Meal Periods — the position on unauthorised extensions of authorised breaks
• Skidmore v. Swift & Co., 323 U.S. 134 (1944) — the origin of the engaged-to-wait distinction

Monitask reports idle time against each person's own schedule, with the threshold visible rather than buried in a default.
Start a free Monitask trial