The single largest cost in litigation is not filing fees, depositions, or outside counsel rates. It is document review — and most of what drives that cost was avoidable if the right decisions had been made three stages earlier.
The invoice arrived and the number was larger than anyone had expected. Not because something went wrong. Not because the case was unusual. Simply because the volume of documents that entered review was larger than it needed to be — and by the time anyone noticed, the review was already running.
This is the most common pattern in eDiscovery cost overruns. According to the American Bar Association, document review accounts for over 80 percent of total litigation spend — roughly $42 billion annually in the United States. According to the RAND Institute for Civil Justice, it costs approximately $18,000 to manage a single gigabyte of data through full review. A legal team that collects 100 gigabytes when 30 would have been sufficient just paid for 70 gigabytes of unnecessary review work. That is not a billing issue. It is a scope control issue — and it begins long before the first document is opened.
The reactive approach to eDiscovery — collect everything, sort it out later — is an extraordinarily expensive default. Organizations that jump into full-scale review without a structured early case assessment process routinely overspend by millions on document review, collect far more data than they ever need, and surface critical facts too late to meaningfully influence legal strategy. The Gartner Legal and Compliance Technology Survey found that legal teams with structured pre-review workflows reduce total eDiscovery spend by 30 to 50 percent compared to those without one.
The math is unforgiving. Every gigabyte of data that enters full review carries a cost. A matter where 200 gigabytes are collected when 60 would have been sufficient does not simply cost a little more — it costs roughly $2.5 million more at RAND's estimated review cost per gigabyte. That excess is not driven by the facts of the case. It is driven by the absence of a process designed to limit scope before collection begins.
Outside counsel operating under pressure to manage risk instinctively over-collect. In-house teams that lack visibility into what data exists over-preserve. Vendors that charge per gigabyte have no financial incentive to narrow scope. The result is a systematic bias toward more data, more cost, and less strategic clarity — and it begins at the identification stage, not at review.
Every gigabyte that enters full review costs roughly $18,000. The question early case assessment answers is how many of those gigabytes actually need to be there.
Broad collection across all custodians and systems. No data profiling before review. Review team surfaces case-critical facts weeks into production. Budget overrun discovered at invoice.
Data profiling and ECA before full collection. Custodian prioritization. Targeted collection narrows scope. Case-critical facts surfaced early enough to influence strategy. Budget controlled.
Early case assessment — or ECA — is the structured process of rapidly gathering, filtering, and analyzing potentially relevant data at the outset of a matter before full-scale collection and review begins. The goal is to evaluate the scope, cost, risk, and merit of the matter early enough that legal strategy can be shaped by evidence rather than assumptions.
In practice, this means profiling data sources before committing to full collection — understanding approximately how much data each custodian holds, what the date range looks like, and whether the data patterns suggest the matter will be large or targeted. It means using analytics tools to identify key concepts, people, and documents in a sample before deciding what to collect in full. And it means having the conversation about proportionality — what the matter is worth, what data is actually needed to resolve it, and what the opposing party is likely to request — before the collection scope is set.
The FRCP Rule 26(f) meet-and-confer conference is one of the most underused cost-control tools in eDiscovery. When legal teams enter that conversation with a preliminary data profile and a proposed scope, they have the ability to negotiate limits that serve both parties. When they enter without one, they typically accept whatever the other side proposes — which is usually broader than necessary.
At Sovereign Discovery, we often have the same conversation with legal teams at the beginning of matters: the instinct is to collect broadly and narrow later. The math argues for the opposite. The gigabytes that never enter review are the cheapest ones in the case — because they cost nothing once they are excluded. The gigabytes that do enter review cost $18,000 each.
Building a structured ECA process is not about limiting discovery — it is about making discovery deliberate. Understanding the data landscape early, prioritizing custodians by relevance, profiling data before committing to full collection, and entering the Rule 26(f) conference with actual numbers: these are the disciplines that separate organizations that control their litigation budgets from those that react to them. The 80-percent problem is real. The solution to it begins before review starts.
The $42 billion spent annually on document review in the United States is not a fixed cost of litigation. A significant portion of it is the cost of decisions made too late — or not made at all — about what data actually needs to be reviewed.
Early case assessment is the most consequential cost-control decision in eDiscovery. It is also one of the most consistently skipped. The organizations that build it into their standard workflow are the ones that stop being surprised by their litigation invoices.