Lay out your project’s phases in working days, skip weekends and holidays, and see whether the schedule lands inside your deadline.
Each phase starts on the next working day after the previous one ends, and its length is measured in working days — weekends and the chosen public-holiday calendar are skipped. The effort factor multiplies every estimate so you can stress-test an optimistic or padded plan.
Buffer is the working days between the projected completion and the deadline; a negative buffer means an overrun. Use it to decide where to cut scope or add people.
Working days. A five-day phase that starts on a Thursday finishes the following Wednesday, skipping the weekend.
It scales all estimates at once. Set 1.25 to add a 25% buffer for risk, or 0.8 to model an aggressive plan.
Yes — pick the US or UK calendar (or none) and the planner shifts phase boundaries around those days automatically.
The Deadline Planner breaks a project into up to six sequential phases, each estimated in working days, and maps them onto a calendar from your start date — skipping weekends and public holidays. Use it whenever you need to know whether a real deadline is achievable, how much buffer you have, or which phase will push the schedule into the red.
The headline shows the projected completion date — the last working day of the final phase. Below it, a status badge (on track / tight / overrun) and a slack figure tell you how many working days separate completion from the deadline. The Gantt chart beneath plots each phase as a colored bar scaled to the full timeline; a red vertical line marks the deadline itself. Phases that end before the red line leave buffer; a bar crossing it signals an overrun.
Project starts 2025-06-02 with three phases — Design (5 days), Build (8 days), Test (3 days) — effort factor 1.0, Saturday/Sunday weekend, no holidays, deadline 2025-06-27.
Design runs 2025-06-02 to 2025-06-06, Build 2025-06-09 to 2025-06-18, Test 2025-06-19 to 2025-06-23. Total working days: 16. Completion date: 2025-06-23. Slack: 4 working days before the deadline. Status: on track.
A working day is any calendar day that is not in your chosen weekend pattern and not in the holiday set for your selected country. The planner skips those days automatically when scheduling each phase.
The effort factor multiplies every phase estimate before scheduling. A factor of 1.2 adds a 20% contingency buffer to each phase, so a 5-day estimate becomes 6 effective days. Use values above 1.0 to model risk or historical over-runs.
Slack is the signed count of working days between the projected completion date and the deadline, excluding the shared endpoint. Positive slack means buffer; negative slack means overrun. A slack of 0 or 1 triggers the "tight" status; a slack of 2 or more is "on track"; negative is "overrun".
Yes. Choose United States or United Kingdom from the country selector to automatically skip federal or bank holidays for the project span. You can also enter custom holiday dates to cover other regions or company-specific closures.
Select "Fri–Sat" from the weekend pattern. The planner supports four patterns: Saturday–Sunday, Friday–Saturday (common in the Middle East), Sunday-only, and none (seven-day weeks).
Each phase is scheduled by advancing a cursor forward from the project start date using the addWorkingDays algorithm in ~/lib/workdays. That function increments the date one calendar day at a time, skipping any day whose UTC weekday number appears in the chosen weekend set or whose ISO date string appears in the pre-computed holiday set. The holiday set is built once by iterating known US federal holidays (OPM list) and UK bank holidays (GOV.UK JSON feed) across the full plausible project span, then merged with any custom dates supplied by the user. Effective days per phase equal the raw estimate multiplied by the effort factor, rounded to the nearest integer with a minimum of one. Slack is computed with countWorkingDays, which performs the same skip logic between two dates and returns the inclusive working-day count; one is subtracted to exclude the shared endpoint, and the sign is negated for overruns.
Spot a translation issue, a calculation issue, or have a suggestion? Let us know.