Between Draft and File, Nobody's Checking

In May 2021, a firm called Bond, Schoeneck & King filed a summary judgment motion in the Eastern District of New York. About 250 lawyers. Solid litigation shop upstate. The brief was fine on the merits. It was well written. It cited the right law.

The court killed it anyway.

Not because the arguments were wrong. Because the attorneys submitted their brief before the opposing side had finished theirs. Judge Amon's standing order requires all motion papers to be fully briefed before anything gets filed on ECF. The court terminated the motion and told them to start over.

That rule lives on page three of a six-page PDF on the court's website. If you know to look for it, it takes about ninety seconds to find. If you don't know it exists, you find out when the judge sends your motion back.

This is not a weird edge case. It is the normal state of federal litigation.

The gap everyone walks past

If you build software for law firms, you probably know the workflow. Someone drafts a document. Someone else reviews it. Then it gets filed with the court. Most legal tech products live at one end or the other. Drafting tools help you write. Filing tools help you submit. Calendar tools tell you when things are due.

But between writing and filing, there is a step that almost nobody has built for: checking whether the document actually complies with the assigned judge's rules.

Today, that check happens in someone's head. A senior associate who has appeared before the judge before. A paralegal who maintains a spreadsheet. A docketing clerk who remembers that Judge Block wants a table of contents on anything over 25 pages. At the largest firms, entire teams exist just to track these requirements across judges and districts.

At most firms, nobody is checking at all.

Same elevator, different universe

Here is what makes this problem genuinely hard, and why it is not something a static rule library can solve.

Take that same summary judgment memo and run it past two judges in the Eastern District of New York. They sit in the same building. They share an elevator on Cadman Plaza East in Brooklyn.

Judge Amon's standing order says you cannot file any motion papers until the motion is fully briefed. All moving papers, opposition, and reply must be served before anything goes on the docket. She also requires a hard-copy courtesy package delivered to chambers: opening brief, opposition, Rule 56.1 statements, and exhibits, in a separate binder, everything double-sided. And you need to have held a pre-motion conference before filing any dispositive motion.

Judge Block, one floor away, does not want courtesy copies at all. But he has a hard 25-page limit on memoranda of law. If your brief is 31 pages, you are six pages over, and you either need to file a formal motion requesting more pages or cut a fifth of your argument.

Same building. Same courthouse. Completely different requirements. One judge wants a tab binder with double-sided printing. The other does not want paper. Your brief is either compliant or it is not, and the answer depends entirely on which name is on the case.

Now multiply that by 670 active federal judges across 94 districts. Each one publishes their own individual rules and practices, typically as a PDF, typically buried on a court website, typically updated without notice. The Federal Rules of Civil Procedure set a baseline. Local rules for each district override parts of it. Individual standing orders override both. Three layers of rules, and the one that actually matters in practice is the one that is hardest to find and track.

What we found in the data

Our team spent time looking at what actually happens when attorneys get this wrong. We pulled docket data from the Eastern and Southern Districts of New York and manually reviewed the entry text to verify our findings.

We did not set out to find the biggest number we could publish. We set out to find numbers we could stand behind. For every major claim, we read the actual docket entries, corrected our own queries when they were over-counting, and threw out results that did not hold up on inspection.

What we found, even after being conservative:

Motions terminated by judges for standing order violations. We verified 46 in the Eastern District alone. Motions that were filed, reached the court, and got killed. Not on the merits. On procedure. The bundling rule that tripped up Bond Schoeneck accounted for 31 of them across 16 different judges. Another 15 were terminated because the attorney filed a dispositive motion without requesting a required pre-motion conference first.

Motions filed just to ask for more pages. We found 7,620 across both districts. These are formal “Motion for Leave to File Excess Pages” filings, where an attorney did not know the judge's page limit, wrote past it, and then had to file a separate motion asking permission to exceed it. Each one means someone drafted a brief without knowing the target length, realized too late, and then created more work for themselves and the court. We spot-checked over a hundred of these. Zero false positives.

Courtesy copy failures. 63 verified cases where chambers never received required hard copies. The judge had to issue an order telling counsel to send what should have been sent automatically.

Broader compliance signals. 5,971 cases across both districts where docket entries mention “Individual Rules” alongside the word “violation.” Not all of these are hard kills. Many are warnings, corrections, or orders directing counsel to re-read the judge's practices. But every one of them is a moment where the court had to stop and deal with a preventable compliance failure.

These two districts handle about 7% of all federal case filings nationally. The rules that caused these failures, bundling, pre-motion conferences, page limits, courtesy copies, are not unique to New York. They exist in some form in virtually every federal district.

Why this is a product problem

There is a natural instinct to treat this as a content problem. Publish the rules. Put them in a database. Let attorneys search for their judge. That approach helps, but it does not solve the core issue, because it still depends on the attorney knowing to look.

The attorney who filed that Bond Schoeneck motion did not lack access to Judge Amon's standing order. It was public. It was on the court's website. The problem is that the attorney had to proactively find it, read six pages of procedural requirements, identify which ones applied to a summary judgment motion, and remember all of them while also doing the substantive legal work. That is a workflow problem.

If you are building a filing product and tracking acceptance rates, this is the part of the funnel that is hardest to see. A filing gets rejected or a motion gets terminated, and it looks like a user error. But it is actually a product gap. The user was never told what the requirements were at the moment it mattered.

The fix is not publishing more PDFs. It is bringing the requirements into the workflow. Surface the controlling rules before the attorney starts writing, not after they try to file. Run a check before ECF submission. Trace every requirement back to the source document so the attorney can verify it. This is how compliance moves from tribal knowledge to product infrastructure.

What this means

The calendar rules market proved that deadline compliance is worth automating. Insurance carriers started requiring it. Malpractice risk drove adoption. Filing compliance is the same category of problem, with the same structural incentive, but nobody has built the infrastructure for it yet.

The teams that get this right will be the ones that treat court rules as structured data, not as reference content. The ones that build the compliance check into the workflow, not as an afterthought. And the ones that understand that in federal litigation, the rules are not just the Federal Rules of Civil Procedure. They are 670 PDFs, written by 670 judges, each one slightly different from the last.