Skip to content
Kairos OneBook a demo
Menu
Advanced · 2 min read

As-planned vs as-built analysis

The retrospective comparison of what was promised against what happened — intuitive to show, hard to prove causation with.

Working definition

As-planned versus as-built is the retrospective family of delay analysis: reconstruct what actually happened (the as-built), set it against the approved plan (the as-planned), and attribute the differences. Its refined form — windows analysis — breaks the project into periods and traces the critical path through each. It is how completed projects get analyzed; its weakness is that showing a difference is not the same as proving what caused it.

Key takeaways

  • Lay the two bar charts on top of each other and the story can look obvious—this started late, that ran long, completion slipped by so many days.
  • Windows (or contemporaneous period) analysis cuts the project into windows — usually update to update — and asks, for each: where was the critical path, what moved it, and who owned that cause.
  • Prospective TIA is commonly considered while work is active; after the event or completion, observational or modeled retrospective methods may be more appropriate.
01

The basic comparison, and why it seduces

Lay the two bar charts on top of each other and the story can look obvious—this started late, that ran long, completion slipped by so many days. The visual is easy to understand, but a raw overlay does not by itself establish responsibility, criticality, or causation. It shows variances, including those absorbed by float, and must be supported by method, records, and path analysis.

02

Windows analysis: the defensible version

Windows (or contemporaneous period) analysis cuts the project into windows — usually update to update — and asks, for each: where was the critical path, what moved it, and who owned that cause. Attribution happens period by period with the schedule that existed at the time, which handles shifting critical paths and concurrent effects far better than one end-to-end overlay. Its cost is data hunger: it needs a credible update for every window, which is precisely what a well-run monthly cycle produces and a poorly-run one cannot reconstruct.

03

Choosing between retrospective and prospective

Prospective TIA is commonly considered while work is active; after the event or completion, observational or modeled retrospective methods may be more appropriate. AACE 29R-03 provides a method taxonomy, and the SCL Protocol advises selecting a method based on the contract, purpose, timing, records, schedule quality, complexity, and proportionality rather than declaring one retrospective method universally correct.

Sources and scope

Follow the reference trail.

This guide is original educational commentary. Kairos reference records identify the source organizations, editions and scope behind this guidance. The original publishers retain their publications and rights.

Recommended Practice 29R-03: Forensic Schedule AnalysisAACE International · checked 2026-08-01Delay and Disruption Protocol, second editionSociety of Construction Law · checked 2026-08-01