Late-Evaluation Rewrite Checklist

Passing the early filter only gets your content retrieved — it still has to survive the generator’s final read, where chunks are dropped silently for being un-self-contained, unattributable, contradictory or redundant. This Google Sheets workbook gives you a section-by-section checklist to rewrite content so it survives that late-evaluation stage. Created by Beatrice Gamba for the AI Search & Agentic SEO course, it operationalises Lesson 3.3’s four late-evaluation tests.

The workbook dedicates one tab to each of the four tests, with a pass/fail/partial scoring guide and the specific fix in each header. T1 — Self-Containment: find every broken reference in a section and replace it with the information it points to. T2 — Attribution: find every statistic and confirm the source, year and scope are all in the same sentence. T3 — Consistency: find claims a competitor also makes and add scope qualifiers to resolve contradictions. T4 — Additive Value: compare your section against a competitor’s and replace overlapping sentences with distinct, more specific claims. Yellow cells mark exactly what to fill in.

Use this for:
‧ Diagnosing why retrieved content still is not being cited (dropped at the generator)
‧ Making each section self-contained, with no broken references
‧ Getting source, year and scope into the same sentence as every statistic
‧ Resolving contradictions with competitor content using scope qualifiers
‧ Replacing overlapping claims with distinct, more specific ones

This is perfect for editors, content teams and SEOs doing the sentence-level rewriting that determines whether a chunk is actually used in an AI answer.