
A review population that arrives at 1.5 million documents is rarely a review problem alone. It is usually the visible result of earlier decisions about collection scope, preservation, processing, and data reduction. To reduce eDiscovery review costs without increasing legal risk, teams need to make those decisions deliberately before document reviewers begin coding.
Attorney review remains one of the most expensive stages of the eDiscovery lifecycle because cost rises with every document that reaches the queue. The goal is not to remove data indiscriminately. It is to identify, preserve, and produce what is responsive while preventing duplicative, irrelevant, or technically unusable material from consuming review time.
1. Start with a defensible collection plan
The fastest way to inflate review spend is to collect every mailbox, shared drive, chat export, and device image available because it feels safer than making scope decisions. Overcollection can create more than cost. It may introduce sensitive personal information, privileged records, and unrelated business data that must later be reviewed, protected, or produced.
Before collection begins, define the relevant custodians, date ranges, systems, issue categories, and likely communication channels. Interview key custodians and information technology personnel early. Their input can reveal whether responsive material lives in Microsoft 365, Google Workspace, Slack, Teams, mobile devices, cloud storage, or a line-of-business application rather than in a traditional email archive.
A targeted plan should still preserve sources that may be relevant. The distinction is important: preservation protects potentially relevant ESI from alteration or loss; collection determines what data enters the active workflow. A certified forensic collection provides the chain of custody and technical documentation needed to defend that distinction later.
2. Cull data before full processing and hosting
Early case assessment is where disciplined teams often find their largest cost reduction. Before paying to process and review an entire dataset, apply agreed-upon filters that remove material outside the matter’s defined scope.
Date filters, custodian filters, file-type exclusions, and domain filters are useful starting points. For example, a commercial dispute may not require system files, executable files, application caches, or family photographs collected from a mobile device. A carefully documented exclusion of those categories reduces processing volume without changing the substantive evidence set.
Search terms can also help, but they should not be treated as a one-time switch. Test terms against sample data, examine false positives, and identify language that key custodians actually used. Broad terms can flood a review set. Narrow terms can omit critical communications. The right approach depends on the claims, the sources, and whether the matter involves technical jargon, code names, abbreviations, or multilingual communications.
3. Use deduplication, email threading, and near-duplicate analysis
Reviewers should not spend hours reading the same content repeatedly. DeNISTing, exact deduplication, email threading, and near-duplicate identification address different forms of repetition, and each has a place in a defensible workflow.
DeNISTing removes common system files that are unlikely to have evidentiary value. Exact deduplication identifies identical files or messages so one representative can be reviewed and coded while duplicate instances remain tracked for family relationships and production needs. Email threading groups related messages and can allow reviewers to focus on the most inclusive email in a conversation rather than reviewing each reply in isolation.
Near-duplicate analysis is especially useful when users circulate slightly revised drafts, presentations, spreadsheets, or form agreements. It does not eliminate the need for legal judgment. A small wording change can be central to a dispute. But grouping substantially similar documents lets reviewers compare the differences quickly instead of encountering those documents randomly across the queue.
These tools must be configured with the matter in mind. Global deduplication can reduce volume dramatically, but teams should consider whether the production protocol, custodian attribution, or opposing-party expectations require a different treatment of duplicates. Cost control works best when technical settings and legal strategy are aligned.
4. Prioritize the documents most likely to matter
Linear review treats every document as equally likely to be responsive. Most matters do not look like that in practice. Analytics can surface communication patterns, recurring concepts, unusual file types, and documents connected to the people or events at issue.
Concept searching, clustering, and communication analysis help review teams understand the data before committing large numbers of documents to manual review. Technology-assisted review can further prioritize likely responsive material based on reviewer decisions. Used appropriately, it can focus attorney time on the records most likely to affect claims, defenses, and investigation findings.
Technology-assisted review is not a substitute for a sound protocol. It requires seed-set decisions, quality-control measures, validation, and a clear record of the process. In a small, straightforward matter, the effort to configure and validate an advanced workflow may exceed the savings. In a large or fast-moving matter, it can be the difference between a manageable review and an avoidable budget problem.
5. Build privilege review into the workflow early
Privilege review is often where a reduced document population becomes expensive again. If privileged communications are not identified until late in the process, teams may need to revisit documents, rerun searches, create logs under deadline pressure, and conduct expensive second-level review.
Use attorney, law firm, and legal-department domains as early indicators, then supplement them with name variations, known counsel lists, and issue-specific terms. Reviewers need clear instructions on the difference between attorney-client privilege, work product, confidentiality, and ordinary business communications involving legal personnel. Those categories are related but not interchangeable.
Privilege analytics and communication mapping can make this work more precise. They can identify likely attorney-client communications and related document families for focused review. They should support attorney judgment, not replace it. A defensible process also includes quality-control checks and a practical escalation path when reviewers encounter uncertain documents.
6. Give reviewers a protocol they can use
A vague coding manual creates inconsistency, rework, and unnecessary escalation. Before the main review begins, establish clear responsiveness definitions, issue tags, confidentiality designations, privilege calls, and family-document rules. Use a small calibration set to make sure reviewers apply those instructions consistently.
The review platform should make those decisions easier, not harder. Saved searches, batch management, standardized layouts, persistent highlights, coding shortcuts, and production-aware fields reduce clicks and prevent avoidable errors. Project managers should monitor coding rates, overturn rates, sampling results, and reviewer questions throughout the review rather than waiting until the final production deadline.
This is also where direct access to experienced support matters. When a search returns an unexpected volume, a native file fails to render, or a proposed production requires special redactions, the answer should come from a team that understands both the data and the legal workflow. Fragmented providers can turn routine adjustments into delays and duplicated effort.
7. Measure cost by stage, not just by total invoice
A low hosting rate does not automatically mean a low-cost matter. Processing charges, user fees, exports, production work, forensic collection, project management, and review labor all affect the actual spend. Ask for reporting that shows data volume and cost at each stage: collected, processed, filtered, hosted, reviewed, and produced.
Those numbers make future decisions more accurate. If one source produces almost no responsive documents, the team can reassess whether similar sources should be expanded. If a search term drives a high rate of nonresponsive hits, it can be refined. If one reviewer group has an unusually high reversal rate, the protocol or training may need attention.
ViewDox helps legal teams connect forensic collection, processing, analytics, review, and production in one managed workflow, so reduction decisions are not lost between separate vendors. That continuity supports faster answers, tighter security controls, and a clearer record of how the evidence was handled.
Reducing cost without cutting corners
The right review strategy is not always the smallest possible document set. A high-stakes regulatory inquiry, a contentious preservation dispute, or a matter involving mobile-device evidence may require broader collection and more intensive quality control. The objective is proportionality: spend review dollars on information that is reasonably likely to matter, and document why exclusions and prioritization decisions were appropriate.
When collection is forensic, culling is tested, analytics are supervised, and review is managed against a clear protocol, lower cost and defensibility reinforce each other. The most useful next step is often simple: examine where documents enter your current workflow and ask whether they truly need to reach a reviewer at all.

Recent Comments