Duration vs Effort: Why Your Estimates Slip
Ask three project managers what a two-day task means and you may get three answers. One means sixteen hours of hands-on work. Another means the bar runs from Monday to Tuesday. A third means the deliverable lands two days later, weekends included. Those are three different measurements, and treating them as one number is the quiet reason estimates slip. This guide pulls effort, duration and elapsed time apart, then puts them back together honestly.
Three Words People Use Interchangeably
Effort, duration and elapsed time are separate quantities that happen to share the same units on the surface. Effort is the raw work content. Duration is the working days the task occupies on your Gantt chart. Elapsed time is the wall-clock span a stakeholder feels when they count calendar squares. When you treat them as one figure, your plan promises dates the team can never hit.
The trap is easy to fall into because everyone answers in days. But a day of work is not the same as a day on the calendar, and neither is the same as a day of waiting. Separate them first, and the rest of estimating gets much easier.
Effort: The Work Itself
Effort is the work content of a task, measured in person-hours or person-days. It answers one question: if someone did nothing but this, hands on keyboard, how long would the doing take? Writing a report might be twelve hours of effort. Painting a room might be six. Effort does not care what day it is, whether the person also sits in meetings, or whether a colleague shares the job.
Effort is the most stable of the three numbers because it is a property of the work, not of the schedule around it. Estimate it first, and estimate it in person-hours, so the calendar cannot contaminate your sense of how big the job actually is. Our guide on how to estimate task durations goes deeper on getting this first number right.
Duration: The Calendar Span
Duration is the number of working days a task occupies, from the day it starts to the day it finishes. This is what the length of a bar represents. Duration is derived, not fundamental. It depends on effort, but also on how many hours per day the person can give, how many people share the load, and how much waiting sits inside the task.
Because duration is derived, it is the number that moves. A twelve-hour task can have a duration of a day and a half if someone works it full time, or two weeks if they touch it an hour a day between other commitments. Same effort, very different bars. That is the whole problem in one sentence.
Elapsed Time: The Wall Clock
Elapsed time is the calendar span from start to finish including everything: weekends, holidays, evenings, and the days the task sat waiting for a reply. It is what a client means when they ask when something will be done and start counting days. Elapsed time is always equal to or longer than duration, because duration counts only working days while elapsed time counts them all.
| Measure | Unit | What changes it |
|---|---|---|
| Effort | Person-hours | Scope and skill |
| Duration | Working days | Availability, allocation, waiting |
| Elapsed time | Calendar days | Duration plus weekends and holidays |
Keep all three in view. You estimate in effort, you schedule in duration, and you promise stakeholders in elapsed time. Trouble starts the moment you use one of these numbers where another belongs, because the gap between them is not small. On a busy team, elapsed time can run two or three times the raw effort, and no one notices until the deadline arrives.
Why the Confusion Wrecks Schedules
The classic mistake is entering effort as duration. Someone estimates a task at three days of work, meaning twenty-four person-hours, and types 3 into the duration column. The Gantt chart now assumes three calendar days of full-time work. But the assignee is on the task half time, loses each morning to meetings, and waits a day for sign-off. The real span is closer to eight working days.
Every task downstream inherits that error, and the project end date drifts. Multiply it across fifty tasks and the schedule is fiction before work begins. A Gantt chart built once and never touched is not a plan, it is a wish. The fix is not heroics later, it is honest conversion up front.
Availability: Nobody Works 100 Percent
Nobody delivers eight hours of task work in an eight-hour day. Meetings, email, breaks, context switching and the ordinary friction of a workplace eat into it. A realistic figure for focused task time is five to six productive hours per day for most knowledge workers, and less again for managers who spend their days unblocking other people.
If you plan at eight hours and the person delivers five, every duration you calculated is off by more than a third before anything goes wrong. Pick a productive-hours-per-day figure your team can actually sustain, write it down, and use the same number in every conversion so your estimates stay consistent across the plan.
Part-Time Allocation Stretches Everything
Most people are not assigned to one task full time. They split their week across projects, support duties and their own backlog. Allocation is the fraction of a person's working day a task can claim. At 50 percent allocation, a task gets half of each day, so its duration roughly doubles relative to full-time work on the same effort.
| Allocation | Productive hrs/day | Duration of 16h effort |
|---|---|---|
| 100% | 6 | 2.7 days |
| 50% | 6 | 5.3 days |
| 25% | 6 | 10.7 days |
Allocation is where estimates quietly collapse. A plan that assumes everyone is fully dedicated is planning for a team that does not exist. Ask what share of each person's time a task truly gets, and let that number stretch the bar.
Multitasking Taxes You Twice
Splitting attention across several active tasks does not just divide time, it adds a tax. Each switch costs re-orientation: reopening files, remembering where you left off, rebuilding the mental context. For people juggling three or more live tasks, that switching cost eats a meaningful slice of the day that never shows up on any timesheet.
The practical result is that someone at 50 percent allocation across two tasks often delivers less than half on each. When you convert effort to duration for a heavily multitasking person, add a buffer for the switching tax rather than assuming a clean division of hours. Better still, reduce how many tasks run at once so the tax shrinks.
The trap is that multitasking feels productive. Touching five things a day looks busy, yet each one crawls because none of them ever gets an uninterrupted run. If your durations keep overshooting even after you account for availability and allocation, an unpriced switching tax is usually the culprit hiding in plain sight.
Waiting: The Silent Duration Eater
Some duration is not work at all. It is waiting: for an approval, a code review, a dependency to finish, a vendor to reply, or paint to dry. Waiting adds nothing to effort but everything to duration and elapsed time. A task with two hours of effort can span a week if it waits five days for sign-off.
Waiting is often invisible in estimates because no one is working, so it feels free. It is not. Model waiting explicitly, either as lag on a dependency or as its own line of duration, so the bar tells the truth. Hidden waiting is one of the most common reasons a plan that looked tight on paper slides week after week.
Converting Effort to Duration Honestly
Converting effort to duration is arithmetic, not guesswork, once you have the right inputs. The formula is simple: duration equals effort divided by productive hours per day times allocation, then add any waiting. Written out as steps, it looks like this.
- Estimate effort in person-hours, ignoring the calendar entirely.
- Set productive hours per day, the focused time the person really gives, often five to six.
- Set allocation, the fraction of each day this task can claim.
- Divide effort by (productive hours per day times allocation) to get working days of active work.
- Add waiting or review time as extra duration, or as lag on the dependency.
- Round up to whole days if your Gantt tool schedules in day units.
Do this for each task and the durations you enter will survive contact with reality. The numbers stop being optimistic guesses and start being defensible forecasts.
A Worked Example: The Report That Slipped
The report that slipped. Priya estimates a market report at 16 person-hours of effort. Her manager types 2 days into the Gantt chart, assuming eight hours a day. But Priya is at 50 percent allocation, gives about six productive hours a day, and needs one day waiting for legal sign-off.
Honest duration = 16 / (6 x 0.5) = 5.33 working days of active work, rounded to 6, plus 1 day waiting = 7 working days. The naive plan said 2 days. The real span is 7. That gap, repeated across a plan, is exactly why projects miss.
Notice that nothing went wrong. No one was lazy and no scope crept. The estimate slipped purely because effort was entered as duration, and the availability, allocation and waiting were never priced in. The math was the whole story.
Putting It on the Gantt Chart
Once your durations are honest, the Gantt chart becomes a forecast you can defend. Longer, realistic bars push out your dependencies and reveal the true critical path. Tasks with room to move show their float, telling you where a slip is harmless and where it threatens the deadline.
Two people, same task. A 40-hour build is assigned first to Sam alone (full time, six productive hours), then in a what-if to Sam plus Dana, both half on it. Solo: 40 / 6 = 6.7, round to 7 days. Together at half each: combined six productive hours a day, still about 7 days, plus coordination overhead.
Adding a person did not shorten it, because effort, not headcount, set the floor. Duration only fell if someone went to full allocation.
Model your real numbers, not your hopes, and you can build the chart in gantts.app in minutes. Adjust allocation, watch durations recompute, and you are planning instead of wishing.
Templates that use this
Keep reading
- How to Estimate Task Durations (Without Guessing)
- Critical Path Method (CPM): Steps, Formula & Example
- Baseline vs Actual: Tracking Schedule Variance on a Gantt Chart
Frequently asked questions
What is the difference between effort and duration?
Effort is the work content of a task in person-hours: how long the doing takes with hands on keyboard. Duration is how many working days the task occupies on your Gantt chart. Duration is derived from effort but stretched by availability, allocation, multitasking and waiting, so the two numbers rarely match.
Why do my tasks take longer than the hours I estimated?
Because you estimated effort but scheduled it as duration. Nobody works eight focused hours a day, few people are on one task full time, and most tasks include waiting for reviews or replies. Availability, part-time allocation and waiting all stretch the calendar span well beyond the raw hours of work.
How do I convert effort into duration?
Divide effort by productive hours per day times allocation, then add waiting. For 16 hours of effort at six productive hours a day and 50 percent allocation: 16 / (6 x 0.5) equals 5.3 working days, plus any review or approval wait. Round up to whole days if your tool schedules by the day.
Does adding more people always shorten duration?
No. Effort sets the floor, and many tasks cannot be split cleanly. Two people at half allocation each give the same combined hours as one full-time person, so the duration barely moves, and coordination overhead can even lengthen it. More headcount helps only when the work genuinely divides and people are actually available.
How much productive time should I plan per day?
Plan five to six focused hours per day for most knowledge workers, and less for managers. Meetings, email, breaks and context switching consume the rest of an eight-hour day. Using eight hours in your conversions overstates capacity by a third or more, which is enough on its own to make an otherwise careful plan slip.