Reviewing, Validating & Documenting the Estimate
An unreviewed estimate is an unexploded error. Estimates are big, detailed, and built under deadline — the perfect breeding ground for omissions, double-counts, and wrong assumptions. Review is the quality gate that turns "we think it's right" into "we've checked it's right." Built on AACE 31R-03.
What estimate review is
This opens the final module of the track, Quality, Validation & Communication. Everything earlier produced the estimate; this module makes sure it's good and that its meaning is conveyed. Review is the first line of that defense — the disciplined check that should sit at the end of every estimate plan (Lesson 16).
The dimensions of a review
A thorough estimate review looks across several dimensions — different lenses that catch different kinds of problems:
- Completeness — is all the scope covered? Are there gaps, missing items, or unpriced elements? (The most common — and costly — failure.)
- Correctness — are the quantities, rates, and arithmetic right? Any double-counts, unit errors, or broken formulas?
- Consistency — do the assumptions, the basis of estimate, and the numbers agree? Does the estimate match its requirements (ERD)?
- Reasonableness — does the bottom line "feel right" against benchmarks, ratios, and experience? Are the headline metrics ($/unit) sane?
- Basis & documentation — is the basis of estimate complete — assumptions, exclusions, contingency all documented (Lesson 14)?
Review methods & documentation
Reviews use a mix of techniques, scaled to the estimate's class and stakes:
| Method | What it catches |
|---|---|
| Checklist review | Standard omissions and required elements |
| Peer / independent review | Errors and blind spots the author missed |
| Benchmark / metric checks | Bottom lines that are out of line with norms |
| Cross-checks (top-down vs bottom-up) | Structural errors in build-up |
Using review well
Build review into the estimate plan with real time allocated, scale its depth to the estimate's class and stakes, prioritize completeness and reasonableness (where the big errors hide), use an independent reviewer for anything significant, and document the review and its resolution. A reviewed-and-documented estimate is one a manager can commit to and a future auditor can trust.
Ten things to remember
- Review checks a completed estimate for completeness, correctness, and reasonableness.
- It's the quality gate before an estimate is issued — opens Module 2D.
- Plan it in with real time — it's the first thing skipped under pressure.
- Check completeness, correctness, consistency, reasonableness, and the basis.
- Completeness errors are the silent killers — missing scope has nothing to flag it.
- Use checklists, peer review, benchmarks, and top-down/bottom-up cross-checks.
- An independent reviewer is the most effective — authors are blind to their own assumptions.
- Document the review — who, what was found, what was resolved.
- Scale review depth to class and stakes — more rigor for bigger decisions.
- Review catches errors; validation (next) confirms fitness — distinct, complementary steps.
Glossary
- Checklist review
- A standard list of items and common omissions.
- Completeness check
- Verifying all scope is covered and priced.
- Cross-check
- Comparing top-down and bottom-up builds.
- Estimate review
- The quality check of a completed estimate before issue.
- Independent reviewer
- A qualified checker who didn't build the estimate.
- Reasonableness check
- Testing the result against benchmarks and experience.
- Review documentation
- The record of who reviewed and what was found.
- Validation
- The separate, formal fitness-for-purpose confirmation.