E-discovery is the process of finding, preserving, and handing over electronically stored information, or ESI, in a lawsuit. In federal court, the frame comes from the Federal Rules of Civil Procedure: Rules 26 and 37 set the duty to preserve and the scope of what changes hands. The work runs from legal hold to production set, and the EDRM model maps the stages.
Each stage has its own job: identify data, preserve it, collect it, process it, review it, and produce it. Skipping a stage is costly. Under Rule 37(e), lost data that should have been kept can bring curative court orders, and worse if a court finds intent. This connects to our earlier piece, How Technology-Assisted Review Actually Works — and What Rule 26 and Rule 502 Require Before You Deploy It.
When does the duty to preserve data start?
The duty starts when litigation is on the horizon. Rule 37(e) covers ESI that should have been preserved in the anticipation or conduct of litigation. The first tool is the legal hold notice, which tells custodians to keep relevant data in place. Once a dispute is reasonably expected, deletion and purge cycles stop for covered data.
The hold notice is a record as much as a command. It lists who was told and when. Courts read that paper when data later goes missing.
What does Rule 26(b)(1) let parties seek?
Per Rule 26(b)(1), parties may seek any nonprivileged matter relevant to a claim or defense and proportional to the needs of the case. The rule lists the factors courts weigh: the issues at stake, the amount in controversy, each side's access to information and its resources, and whether the burden outweighs the benefit.
Scope gets set early. The Rule 26(f) conference brings counsel together to form a discovery plan. Initial disclosures follow within 14 days of that conference, per the rule. The plan may also settle the forms in which ESI should be produced.
What happens during collection and processing?
Collection is the transfer of data from the company to counsel, per the standard EDRM account. Forensic care matters here, because mishandling can alter metadata, the timestamps and details that ride along with files. Processing then prepares native files for a review platform, with text and metadata pulled out and early culling steps such as dedupe applied.
Size is decided upstream. The identification stage sets the scale of the collection, so a loose scope there ripples through every later step. Care at this stage also protects the files themselves.
How does review end in production?
Review filters the set for relevance and privilege before anything moves. Production then turns documents over to the other side under agreed specs, often with a load file for the receiving platform, per the EDRM account of the stage. Files may go as native versions or as fixed formats such as PDF or TIFF.
Sanctions risk sits behind this work. Under Rule 37(e), a court may order measures to cure prejudice from lost ESI. Only on a finding of intent to deprive may it presume the data was unfavorable, instruct the jury, dismiss the case, or enter default.
What does this mean in practice?
- Issue the written hold as soon as a dispute looms, and log who got it.
- Stop deletion schedules for covered custodians and data sources.
- Track chain of custody from collection through processing.
- Match the production format to the specs the parties agreed on.
Each step leaves a paper trail. That record is what a court reviews if data goes missing. Courts tend to reward the team that can show its steps.
Conclusion: Order Carries the Process
From hold notice to production set, e-discovery runs on sequence and records. Rules 26 and 37 supply the legal frame, and the EDRM model supplies the map. Teams that document each stage move data without surprises. This article is for informational purposes only and is not legal advice — for decisions about specific facts, work with qualified counsel. Readers following this should also see AI Meeting Note-Takers and Privilege: Closing the Governance Gap Legal Teams Keep Discovering.
