12 Discovery Stipulation Mistakes and How to Avoid Them

This commercial dispute with a former client went to litigation over issues with their server logs some years ago. We agreed to preserve and produce all server logs relating to their account. Simple enough you would think. No, it wasn't. The VPS instances for that client were on shared hypervisor nodes with dozens of other customers. Therefore all the activity on those instances was mixed up together in the server logs for all of them. In the end we had to keep log rotation for all production servers on all of the nodes on all of the instances on all of the servers, on all of the campuses, forever (or at least for several years), on several terabytes of storage that we had no budget for. Every production file had to be fully redacted to prevent disclosure of trading information of other clients. The cost of the litigation to resolve the actual commercial dispute was less than the cost of the discovery to produce the server logs.

This is how lawyers end up in trouble with "relating to". All a lawyer has to do is read his/her side of the contract and he/she has declared that all emails relating to the contract must be produced. The contract does not say anything about "account". Logs are kept by host, service, etc. with time stamp, not by "account".

All stipulations today get signed off by the engineer(s) who created the data map for said stipulation (i.e. specific systems, specific log types, specific fields, within a specific date range, and exported in a specific fashion). I require a sample production first (e.g. production of 1 week's worth of data) so that party can see what it is that they are actually requesting prior to agreeing to production of 6 months of data, etc. Furthermore, a Protective Order should be sought to protect the data of third-party customers and a cost-sharing clause to ensure that request for data that resides on shared infrastructure does not become cost-prohibitive.

Stipulations that read as cooperative on the surface can prove very expensive to implement once someone is tasked with implementing them.

A discovery issue I would approach differently is agreeing to overly broad language about producing electronically stored information without defining what the parties mean by a reasonable search. Everyone may leave the negotiation thinking they have reached a practical compromise, only to disagree later about search terms, custodians, or how much material each side must review. That ambiguity can create more disputes than the stipulation was supposed to prevent.

Today, I would negotiate the mechanics alongside the general obligation. I would identify the relevant sources, discuss search methods, establish a process for resolving disputes, and clarify how the parties will address unexpected costs. A stipulation should eliminate uncertainty, not simply postpone the argument.

One area where I would be cautious is agreeing to a discovery schedule without considering how difficult it may be for a spouse to obtain records from outside sources. Family law cases often involve tax returns, business valuations, retirement accounts, and financial institutions that operate on their own timelines. A deadline that looks reasonable to the attorneys may leave a client unable to obtain important evidence through no fault of their own.

I would approach that negotiation by identifying the records that matter most, estimating how long third party requests might take, and building in a defined procedure for addressing delays. I would rather resolve those practical issues at the beginning than spend time later arguing over extensions. A well drafted stipulation should make the case easier to manage, not create a procedural trap.

I think the stipulation that hurt us the most was agreeing to produce all communications with the buyer in our commission dispute from my real estate practice. It sounded fair or even generous when agreed to, but nearly all of our communications with the buyer were done via WhatsApp on agents' personal phones. Many of those communications were embedded in agents' family chats. Often the communications were done via voice notes or even included messages with other clients with no connection to the case at hand.

Initially, I hadn't realized the volume of communication, but the human impact has been much greater. Some agents have been reluctant to send messages to clients via their personal phones for fear of those communications being discovered. Two of our long-standing clients have been concerned that we have been reviewing their communications with other staff. We didn't lose the dispute, but we did lose some internal trust and some external goodwill that is difficult to recover.

I would now approach this negotiation as I would a viewing of a house prior to entering into a contract to purchase. I would have the individuals who actually communicated with the buyer explain to me how they communicated and where they communicated. I would look to agree named custodians for specified channels over a tight time period, with scope to abide by any necessary restrictions in relation to communication with any other buyers. I would ensure that all team members are aware of the process and that there is no surprise to them.

Terms and conditions of discovery are written by lawyers but implemented by your staff. Therefore, blanket agreement to terms that appear to be cooperative are only as good as your team's understanding of how your business really communicates on a daily basis.

In family law, I would be particularly careful about stipulations that restrict access to financial records before both sides have a clear picture of the marital estate. An agreement that limits the time period for producing bank statements or business records might seem reasonable at first, but later disclosures can reveal transactions that require a closer look. A narrow agreement can become a real obstacle if it prevents a party from tracing assets or understanding the other spouse's financial position.

Today, I would build a process into the stipulation for requesting additional records when the initial production raises legitimate questions. I would want the parties to define the original limits clearly while preserving a practical way to address new information. That balance can keep discovery focused without compromising the financial information needed to resolve the case fairly.

I would be wary of a stipulation that allows the parties to exchange documents in a convenient format without specifying what that format must preserve. A PDF may be adequate for some records, but spreadsheets, databases, and other electronically stored information may lose important functionality or context when converted. The parties can technically comply with the agreement and still end up arguing over whether the production is usable.

Today, I would address file formats, metadata, and searchable text during the negotiation rather than after production begins. I would tailor those requirements to the types of records involved instead of demanding every available data field. A little precision upfront can save substantial time and expense later.

I would be cautious about agreeing to a discovery deadline without accounting for the steps that must happen before it. Producing documents, reviewing them for privilege, arranging depositions, and obtaining information from third parties all take time. A schedule that looks manageable on paper can become unrealistic once the parties encounter delays or disputes.

Today, I would work backward from the key deadlines and identify the dependencies before agreeing to dates. I would also address extensions in the stipulation, including whether the parties can agree to limited adjustments or need court approval. That approach helps prevent a seemingly minor scheduling problem from interfering with substantive preparation.

When broad restrictions override the specifics involved in the creation of all documents, for example, proofs, drafts, and the whole internal auditing process, this results in a great deal of complication instead of clarity in terms of the law. I had a chance to participate in several business projects that required the execution of all the drafts stored in a particular contract management system. The idea seemed to be good in terms of transparency but caused significant problems at the stage of technical implementation. The enormous number of repetitive metadata and autosaved drafts led to the failure of the review system, which resulted in delays of the process along with expenditures connected to it. Moreover, the production of various complicated documents in their original form creates security issues since they include external information that may be important for the business but is not connected to the needs of the investigation itself.

If I had faced this situation again, I would change the strategy and drive at the tiered discovery protocol according to which we would first provide the final agreement and the fundamental verification process. Instead of the requirement regarding the format of the document, I would state the necessity of having a PDF format with the limited number of metadata, such as signer's ID, IP address, and the time the document was drafted, thus making it possible to collect the pieces of evidence without being overwhelmed with unnecessary documents. The production of the original files should be reserved for the cases when the inner logic of documents and some data could be important for the case. It ensures the effectiveness of the discovery process while meeting the requirements laid down by modern business document systems.

A prior stipulation treated shared drives as a single source, without distinguishing active folders from archived, inherited, or access-restricted material. That decision created confusion when files had multiple owners, outdated permissions, and unclear version histories. The legal issue became tangled with ordinary recordkeeping weaknesses.

Today, I would separate repositories by governance status before negotiating scope. The stipulation should identify who controls each environment, how versions are preserved, whether deleted files are recoverable, and how access logs will be handled. I would also seek agreement on family relationships between files and attachments. In large organizations, information architecture reflects years of growth, acquisitions, and workflow changes. Discovery terms must recognize that infrastructure is rarely as tidy as an organizational chart suggests.

A stipulation regarding search terms appeared balanced because both sides exchanged proposed keywords. Its weakness was the absence of a testing process. Several broad terms produced large volumes of irrelevant material, while narrower terms missed informal language employees actually used. I saw that negotiated words can become a substitute for understanding the business vocabulary at issue.

Today, the process would begin with a small sample set and documented hit reports. The parties could assess false positives, abbreviations, and communication patterns before applying terms across the full collection. Search methodology should be iterative, with reasonable transparency, rather than a one-time bargain that locks both sides into flawed assumptions.

I have seen broad confidentiality stipulations create unnecessary friction when they define nearly every produced document as confidential. In injury litigation, records often need to be reviewed by treating providers, consultants, and family members helping an injured person understand the claim. A vague restriction can turn ordinary preparation into a series of permission requests and create avoidable disputes over who may see what.

A revised stipulation should require designation standards, a prompt challenge procedure, and language allowing disclosure to identified medical and litigation participants who agree to the order. Sensitive information deserves protection, but blanket secrecy impedes development. It should protect privacy without blocking meaningful review of causation, damages, or liability.