Chain three or more updates and see when the finish date actually moved.
What it does
Takes a series of schedule updates, orders them by their data dates, and compares each neighbouring pair. Each pair is a 'window'. You see how the project finish date drifted across the whole chain, which single window moved it the most, and exactly what changed inside every window.
The project finish date across a chain of updates: each step is one window.
What it helps you achieve
Attribute delay to a specific period instead of a whole project.
Find the window where the finish date slipped and drill straight into the changes behind it.
Build a period-by-period narrative that matches how delay is normally argued.
How to run it
Where to find it: Project page → Windows Analysis
Upload three or more schedule updates to a project (up to seven per analysis).
Open the project and choose Windows Analysis, then + New windows analysis.
Tick the updates to include, in any order: they are always chained chronologically by their data dates.
Set the Near-critical threshold (working days) and choose Run Analysis.
Read the headline: total finish-date drift, the number of windows, and the largest single-window shift.
Open any window to see the full comparison for that period, or choose Download Excel.
What you get
A chart of the project finish date across the chain.
The finish-date shift for every window, with the largest one flagged.
The complete comparison for each window on demand.
A saved result you can reopen from the history list.
Optionally, a modeled delay at the start of any window (see Time Impact Analysis).
Good to know
Windows are compared using the dates stored in each update, exactly as P6 reported them.
Re-running the same chain with the same settings reuses the saved result.
Limits, stated plainly
It needs at least three updates with data dates; a file without a data date can't be placed in the chain.
It shows when and what moved, not who is responsible: causation and entitlement remain a judgment for the analyst.
Try Windows Analysis on your own schedules. Claims Digger is invite-only while we’re in early access.
Between three and seven in one analysis. Each neighbouring pair becomes one window.
Is this the same as a Time Impact Analysis?
No. Windows Analysis measures what happened between real updates; a Time Impact Analysis models the effect of a hypothetical delay. You can combine them: model a delay at the start of any window and compare it with what the window actually did.