> Short answer: A project management tool with built-in time off tracking keeps
leave requests, public holidays and working patterns in the same system that holds
the work. That matters because capacity is a subtraction problem: you cannot know
what a team can finish next week without knowing who is absent. When leave lives in
separate HR software, the project tool confidently plans work for people who will
not be there.
Why almost no tool does both
The software market split this problem in two a long time ago, and the split has nothing to do with how teams actually work.
HR tools own leave. They are built around policy, compliance and payroll: accrual rules, statutory minimums, carryover limits, approval chains. They know Ayşe has 14 days and has used 6. They have no idea what Ayşe is supposed to deliver on Thursday.
Project tools own work. They own tasks, boards, deadlines, dependencies. They know Ayşe owns three tasks due next week. They have no idea she is in Antalya.
Both systems are individually correct and jointly useless, because the question a team lead actually asks sits exactly on the seam: *given who is actually available, what can we realistically finish?*
The usual workaround is a third artifact — a shared calendar, a pinned Slack message, a spreadsheet tab called "who's out." It works for about two months, then someone forgets to update it, and the planning error it produces is invisible until a deadline is missed.
What breaks, concretely
These are the failure modes worth recognising, because each one looks like a different problem on the surface.
- Capacity math is silently inflated. Any estimate of "the team has 200 hours next
- Deadlines land on absent people. A task assigned Tuesday, due Friday, to someone
- Handovers don't happen. Leave approved in an HR tool produces no signal in the
- Public holidays are forgotten across borders. A distributed team plans a sprint
- The "who's out?" tax. Every week, someone asks in chat. Every week, three people
The capacity subtraction, worked
Here is the arithmetic that a leave-aware tool does for you and a leave-blind one cannot. Take a five-person team, a normal week.
Start with theoretical capacity:
| | | |---|---| | 5 people × 5 days | 25 person-days |
Now subtract what is actually unavailable:
| Subtraction | Days |
|---|---|
| One national holiday (everyone) | −5 |
| One person on annual leave, 3 days | −3 |
| One person on a half-day | −0.5 |
| Remaining calendar days | 16.5 |
The team that planned against 25 will end the week at about 45% of its plan and call it an execution problem. It was an arithmetic problem, and the inputs for the correct arithmetic were sitting in a different browser tab.
The number itself matters less than the habit: capacity is what remains after subtraction, and the subtractions have to come from somewhere the plan can see.
The three usual workarounds, and what each one costs
Before buying anything, it is worth being precise about what you already have, because most teams are running one of these three and have never named the cost.
| Workaround | What it gets right | What it costs you |
|---|---|---|
| Shared calendar (Google/Outlook "Team Absence") | Free, everyone already has it, absence is visible at a glance | Nothing subtracts. No balances, no approval, no record. Someone can book leave they don't have and nobody finds out until year-end |
| Spreadsheet ("who's out" tab) | Flexible, costs nothing, you can model any policy | Accuracy decays the moment one person forgets. No permissions — anyone can edit anyone's balance. Not an employment record you'd want to defend |
| HR tool + project tool, unlinked | Both halves are individually correct and compliant | The seam is manual. Approved leave produces no signal in the work tool, so capacity views stay wrong and the integration is a person remembering to check |
The case for a leave-aware project tool is not that the other options are wrong. It is that all three ask a human to be the integration layer, and humans are reliably bad at that specific job.
Distributed teams make this sharper
Everything above gets harder the moment a team spans borders, and this is where the gap stops being an annoyance and starts being expensive.
A team split across Germany, Turkey and the UK is running three different holiday calendars with almost no overlap. German public holidays vary by federal state. Turkey has two religious holidays whose dates move each year on the Gregorian calendar and which extend into multi-day national closures. The UK has separate bank holiday sets for England/Wales, Scotland and Northern Ireland.
None of that is exotic — it is just three ordinary countries. But it means the question "is everyone available on the 23rd?" cannot be answered from memory by anyone, and a tool that stores holidays as free-text tasks will be wrong within a quarter.
This is also the argument for country presets over manual entry. Typing one country's holidays is tedious; typing three and keeping them current for years is a task that quietly stops happening.
What "built-in time off tracking" should actually include
Not every tool that shows a vacation badge is doing this properly. This is the checklist I would test against, in rough order of how often each one turns out to be missing.
Working pattern
- Configurable workweek — a team where the weekend is Friday–Saturday is not an edge case
- Half-day granularity, because most leave is not in whole weeks
- Per-person employment start date, which accrual depends on
- Multiple leave types with different rules — annual, sick, personal, parental, unpaid
- Per-type settings: does it need a reason, an attachment, advance notice, approval?
- Accrual mode: does the balance reset annually, or accrue from the employment
- Carryover: how many days roll over, and do they expire?
- Balance enforcement: can someone book days they do not have — blocked, or warned?
- Public holidays as first-class objects, not tasks someone typed in
- Recurring rules, so you configure Christmas once and not every year
- Country presets, because typing 13 national holidays by hand is where this fails
- Approval chain — who approves, and does a one-day request need the same path as a
- An audit log. Leave balances are an employment record; "who changed this and when"
The integration itself
- Approved leave appears in the same calendar as the work
- Capacity and workload views subtract it automatically
- Assigning a task to someone who is away surfaces it at assignment time, which is
If a tool covers the first three groups but not the last, you have bought an HR module that happens to live in a project tool. The integration is the entire point.
How to test a tool in fifteen minutes
Vendor feature lists will claim all of the above. These four checks separate the tools that mean it:
- Book leave, then open the capacity view. Did the number go down on its own? If
- Import a country's public holidays. If there is no preset and you are typing
- Assign a task to someone who is away. Does the tool tell you at that moment, or
- Change someone's balance, then look for the record. No audit trail means you
Where Poitim fits — and where it doesn't
Poitim was built with this seam closed, because we kept hitting it ourselves running agency work at poi369.
What is in the product today: a configurable workweek and half-day granularity; seven leave types (annual, sick, personal, maternity, paternity, unpaid, other) each with its own rules for reason, attachment, notice period and approval path; two accrual modes — annual reset and employment-anniversary — with configurable carryover and expiry; balance enforcement that can block or warn; workspace-level public holidays with recurring rules; country presets for Germany, the Netherlands, Turkey, the UK and the US; and an audit log on balance changes.
The anniversary mode implements seniority tiers where the law expects them — under Turkish labour law, annual entitlement steps from 14 days to 20 at five years and 26 at fifteen — with first-year entitlement pro-rated by months since hire rather than the stricter statutory zero.
Where it doesn't fit. Poitim is not an HR system. There is no payroll, no contract management, no performance review cycle. If you need those, you need an HR platform and Poitim is the wrong shape. Five country presets is a start, not global coverage — every other country means configuring the policy by hand, which is supported but is work. And Poitim prices per team rather than per seat, with a member cap on each plan (2, 5, 15 and 40 on the four tiers) — that is cheaper than per-seat pricing for most teams, but it is not unlimited, and if your team is 60 people this is not your tool yet.
Frequently asked questions
Do I need a separate HR tool if my project tool tracks leave? It depends on what else you need. Leave tracking, holidays and approvals are what overlap with planning. Payroll, contracts, performance reviews and recruitment do not touch the project plan and belong in an HR platform. Small teams often find the leave half is the only half they needed.
What is the difference between annual reset and anniversary accrual? Annual reset gives everyone their full allowance on a fixed date, usually 1 January. Anniversary accrual ties entitlement to each person's employment start date, so somebody who joined in September gets their reset in September. Some jurisdictions require the anniversary model, and seniority tiers only make sense under it.
Why do public holidays need country presets? Because a distributed team spans several holiday calendars and entering them by hand is both tedious and easy to get wrong. A preset seeds the year's holidays for a country in one action; recurring rules mean fixed-date holidays configure themselves in future years.
Can a shared calendar do this instead? It can display absence. It cannot subtract it from capacity, block a booking against an insufficient balance, route an approval, or leave an audit record. A calendar answers "who is out"; it does not answer "what can we commit to."
How does leave affect sprint planning? It changes the size of the sprint. A two-week sprint with a public holiday and one person on a week's leave has roughly 20% less capacity than the previous one. Teams that don't adjust carry the difference as unfinished work and read it as a velocity decline.
What should the audit log record? Balance changes, manual adjustments, who made them and when, and the policy in effect at the time. Leave balances are an employment record, and the question "why does this person have 17 days" needs an answer that is not somebody's memory.