E-Discovery Collection & Review Summary
Internal status report tracking ESI from collection through attorney review, with completion projections.
Prerequisites
- Custodian list with collection scope (systems, date ranges, locations)
- Collection logs and processing reports from vendor
- Review platform exports (coding, reviewer activity, search-term results)
- Search-term list or Boolean queries
- Review protocol (relevance, privilege, issue tags) and QC standards
- Key deadlines the review must meet
- Protective order or confidentiality requirements for reporting
Quick Start
- Gather source data (collection logs, processing reports, review exports, search-term results, privilege summaries).
- Normalize custodian names into a master mapping table.
- Run integrity checks — pipeline totals must satisfy: Collected ≥ Processed ≥ Loaded ≥ Reviewed.
- Compute core metrics tables (see below).
- Draft executive summary with risks and recommendations.
- Apply work-product legend.
Core Metrics
A. Collection & Processing by Custodian
Per custodian: Collected GB, Item Count, Processed %, Load Failures, File-Type Mix, Exceptions.
B. Search-Term Effectiveness
Per term/query: Hits, Hit Rate %, Sample Precision %, Overlap %, Recommended Action.
C. Review Status (Matter-Level)
Total Reviewable, Reviewed Count, % Reviewed, Relevant %, Privileged %, Non-Responsive %, Est. Completion Date.
D. Reviewer Productivity
Per reviewer/team: Docs Reviewed, Hours, Docs/Hour, Relevant %, Privileged %, QC Rate %.
E. Privilege Patterns
Per custodian: Privileged %, Dominant Doc Types, Red Flags.
Key Formulas
| Metric |
Formula |
| Hit Rate % |
hits ÷ total reviewable corpus |
| Precision % |
relevant hits ÷ sampled hits |
| Review Velocity |
docs reviewed ÷ time period |
| Completion ETA |
remaining docs ÷ current velocity |
| Privilege Rate % |
privileged ÷ reviewed |
Executive Summary Template
- Collection completeness — status and anomalies
- Processing health — success rate and key errors
- Search-term performance — top/weak terms and overlap
- Review progress — reviewed %, relevance %, privilege %
- Forecast — ETA with assumptions
- Risks & actions — top three items
Integrity Checks
- Custodian names align across all sources (master mapping table).
- Pipeline totals are monotonically decreasing through each stage.
- Every search term in the term list appears in the results export.
- Coding records contain reviewer, date, relevance, and privilege fields.
- Processing errors are reconciled against re-runs.
Findings & Recommendations
- List missing sources and resulting limitations.
- Flag custodians with outlier volumes or error rates.
- Identify low-precision or high-overlap search terms.
- Note reviewer inconsistencies and QC actions taken.
- Recommend resource changes or search-term refinement.
- Document assumptions and gaps with impact and mitigation.
Pitfalls
- Report only aggregate metrics — never disclose privileged content.
- If counts conflict, surface the discrepancy and state the suspected cause.
- Use U.S. discovery terminology; confirm local rules before external distribution.
- Mark uncertain legal citations with
[VERIFY].
- Keep narrative neutral and data-driven; avoid strategy conclusions.
- Always include: "Attorney Work Product / Privileged & Confidential. Internal use only."
1---2name: ediscovery-review-summary3description: Produces an internal U.S. litigation e-discovery status summary covering ESI collection, processing, search-term hits, review throughput, relevance and privilege rates, and completion forecasts. Triggers when the user requests an e-discovery status report, collection log summary, processing report, review statistics, coding report, search-term results, or privilege review summary.4license: Apache-2.05---67# E-Discovery Collection & Review Summary89Internal status report tracking ESI from collection through attorney review, with completion projections.1011## Prerequisites1213- Custodian list with collection scope (systems, date ranges, locations)14- Collection logs and processing reports from vendor15- Review platform exports (coding, reviewer activity, search-term results)16- Search-term list or Boolean queries17- Review protocol (relevance, privilege, issue tags) and QC standards18- Key deadlines the review must meet19- Protective order or confidentiality requirements for reporting2021## Quick Start22231. Gather source data (collection logs, processing reports, review exports, search-term results, privilege summaries).242. Normalize custodian names into a master mapping table.253. Run integrity checks — pipeline totals must satisfy: Collected ≥ Processed ≥ Loaded ≥ Reviewed.264. Compute core metrics tables (see below).275. Draft executive summary with risks and recommendations.286. Apply work-product legend.2930## Core Metrics3132### A. Collection & Processing by Custodian3334Per custodian: Collected GB, Item Count, Processed %, Load Failures, File-Type Mix, Exceptions.3536### B. Search-Term Effectiveness3738Per term/query: Hits, Hit Rate %, Sample Precision %, Overlap %, Recommended Action.3940### C. Review Status (Matter-Level)4142Total Reviewable, Reviewed Count, % Reviewed, Relevant %, Privileged %, Non-Responsive %, Est. Completion Date.4344### D. Reviewer Productivity4546Per reviewer/team: Docs Reviewed, Hours, Docs/Hour, Relevant %, Privileged %, QC Rate %.4748### E. Privilege Patterns4950Per custodian: Privileged %, Dominant Doc Types, Red Flags.5152## Key Formulas5354| Metric | Formula |55|---|---|56| Hit Rate % | hits ÷ total reviewable corpus |57| Precision % | relevant hits ÷ sampled hits |58| Review Velocity | docs reviewed ÷ time period |59| Completion ETA | remaining docs ÷ current velocity |60| Privilege Rate % | privileged ÷ reviewed |6162## Executive Summary Template6364- **Collection completeness** — status and anomalies65- **Processing health** — success rate and key errors66- **Search-term performance** — top/weak terms and overlap67- **Review progress** — reviewed %, relevance %, privilege %68- **Forecast** — ETA with assumptions69- **Risks & actions** — top three items7071## Integrity Checks7273- Custodian names align across all sources (master mapping table).74- Pipeline totals are monotonically decreasing through each stage.75- Every search term in the term list appears in the results export.76- Coding records contain reviewer, date, relevance, and privilege fields.77- Processing errors are reconciled against re-runs.7879## Findings & Recommendations8081- List missing sources and resulting limitations.82- Flag custodians with outlier volumes or error rates.83- Identify low-precision or high-overlap search terms.84- Note reviewer inconsistencies and QC actions taken.85- Recommend resource changes or search-term refinement.86- Document assumptions and gaps with impact and mitigation.8788## Pitfalls8990- Report only aggregate metrics — never disclose privileged content.91- If counts conflict, surface the discrepancy and state the suspected cause.92- Use U.S. discovery terminology; confirm local rules before external distribution.93- Mark uncertain legal citations with `[VERIFY]`.94- Keep narrative neutral and data-driven; avoid strategy conclusions.95- Always include: "Attorney Work Product / Privileged & Confidential. Internal use only."9697---
ThomasMoreAI