Getting Email Into the Matter File Without Anyone Remembering To

Open any matter file at any firm and it will look reasonably complete. The pleadings are there. The signed agreement is there. The court orders, the expert reports, the executed engagement letter, all present and correctly stored. Then ask the question that matters, which is what the client was told about settlement on the Tuesday before the mediation, and who agreed to the extension, and whether anyone warned the client that the limitation date was tight. Those answers are not in the file. They are in email, scattered across three inboxes, two of which belong to people who are on holiday.

This is the gap that almost every firm lives with and almost no firm has named out loud. The documents are managed. The correspondence, which is where the advice actually happened, is not. Email became the medium of legal work about twenty years ago and firms responded by treating it as a communication channel rather than as the primary record it had quietly become. The filing habits never caught up, because email arrives faster than anyone can classify it and the cost of not classifying it does not land until much later.

What follows is about the mechanics of closing that gap. Not a plea to be more disciplined, because discipline is not the constraint here, but a look at why filing after the fact structurally fails, why filing at the moment of sending works, what genuinely belongs in a matter file and what does not, and why a file that lives inside one person's inbox is a risk the firm carries whether or not it has noticed.

The Record of a Matter Is Mostly Correspondence

Think about what a completed matter actually consists of. There will be a modest number of formal documents, perhaps thirty or three hundred depending on the practice area, and every one of them is the product of a much larger volume of correspondence that shaped it. The advice, the instructions, the concessions, the client's changing risk appetite, the moment where a party said something they later denied saying, all of that lives in email. If you removed the emails and kept the documents, you would still have the outcome but you would have lost the reasoning entirely.

This matters practically rather than philosophically. When a fee is disputed, the question is what the client was told about scope and when. When a claim is threatened against the firm, the question is what advice was given and whether the risk was flagged. When a matter is reassigned, the question is what the incoming lawyer needs to know that is not obvious from the pleadings. Every one of those questions is answered by correspondence, and every one of them is asked at a moment when nobody has time to reconstruct a chronology out of a search across several people's mailboxes. The file either holds the answer or it does not.

Why "I Will File It Later" Is Not a Character Flaw

The standard theory of email filing is that lawyers should drag important messages into the matter folder, and that failure to do so is a discipline problem solved by reminding people. That theory has been tested at thousands of firms and it loses every time, which suggests the theory is wrong rather than the people. The reason is that filing after the fact requires a second pass over information you have already processed, and the second pass has no immediate payoff, so it always ranks below anything with a deadline attached.

Watch what actually happens. A lawyer reads a message, decides what to do about it, does that thing, and moves on. The decision about whether it belongs in the file is a separate act of judgement performed on something the brain has already marked as handled. Doing that for eighty messages at the end of a day means re-reading eighty messages, and each one now needs a matter identified, a folder found, and a name chosen. The effort is real and the reward is entirely deferred, so it gets postponed to the weekend, and the weekend version is worse because by then the context has faded and the borderline messages get skipped. Six months of that produces a file that contains the obviously important correspondence and none of the correspondence that turns out to matter.

Filing at the Point of Sending Changes the Economics

The alternative is to make filing happen inside the act of sending, not after it. If the matter is chosen at the moment the message is composed, the classification decision is free, because the lawyer has already decided which matter they are working on. They are not being asked to remember anything or to perform a second pass. They are being asked to name the thing they are already thinking about, in the same breath as writing the message, which is the only point in the lifecycle where the context is fully loaded and the cost is close to zero.

This is why sending from inside the matter, rather than sending from the mail client and filing afterwards, is the single change that fixes the most. The message lands in the client's inbox exactly as it always did, and it lands in the matter record at the same instant, because it was never outside it. There is nothing to remember and nothing to sweep up. The same principle applies to inbound mail: the aim is to capture at the moment of first handling, not at the end of a batch, because first handling is when someone actually knows what the message is about.

The Inbound Half Is Harder Than the Outbound Half

Outbound is solvable because you control the moment of creation. Inbound is messier, because messages arrive at whatever address the sender happens to have, at whatever hour they choose, often on a phone, and frequently from someone who has no idea which matter they are writing about. A client who has three matters open will reply to the most recent thread regardless of subject. Opposing counsel will change the subject line halfway through a chain. A referral source will introduce a new client inside a reply to an old one.

The workable approach is to reduce the volume of email that needs classifying rather than getting better at classifying it. Routine client updates, document delivery, and signature requests do not need to be email at all, and every one you move out of email is one you never have to file. A client portal that is privilege-filtered per document means the client sees what they should see without anyone attaching anything to anything, and because e-signature happens inside the same login with no separate account to create, the entire signature chain stops being a twelve message thread that someone has to reconstruct later. What is left in the inbox after that is genuine correspondence, which is a much smaller and much more filable population.

What Belongs in the Matter File and What Does Not

Not everything that mentions a matter belongs in its file, and firms that try to capture everything end up with a record so noisy that nobody reads it, which is a different way of losing the information. The test worth applying is whether the message would change a reader's understanding of what happened, what was advised, or what was agreed. Substantive advice belongs. Client instructions belong. Anything that fixes a date, concedes a point, changes scope, or records a decision belongs. Anything that documents a warning given, especially about risk or cost, belongs twice over.

The other side of that test is equally important. Scheduling noise, calendar confirmations, delivery receipts, and internal chatter about who is covering the hearing do not belong in the matter file, and neither does candid internal commentary about the client, opposing counsel, or the judge. Assume that the matter file may one day be read by the client, by an insurer, or in a production set, and write into it accordingly. That is not about hiding anything. It is about recognising that a matter file is a professional record with an audience, and that the place for internal working notes is a channel that is understood by everyone to be internal.

Attachments Are Documents, Even When They Arrive as Email

An email with a draft attached is two records, not one, and firms routinely capture only the first. Filing the message preserves the covering note and the timestamp. It does not put the draft into the document set, which means the operative version of an agreement can end up existing only as an attachment inside a thread, discoverable only by someone who thinks to search the mail rather than the documents. That is how a firm ends up with three versions in circulation and no authoritative answer about which one was sent.

The fix is to treat the attachment as a document the moment it arrives, stored where documents live rather than where messages live. In Casely, documents sit under AES-256 encryption with a per-firm key, and every document carries a comment field recording what changed and why, so the story of a draft is written in language a colleague can read rather than inferred from the order of emails in a thread. The covering email still belongs in the correspondence record, because the note that accompanied the draft is often the most legally significant part of the exchange, but the draft itself needs to leave the inbox and join the document set.

A File That Lives in One Inbox Is a Risk the Firm Owns

Here is the part that tends to get noticed only when it is too late. If the substantive record of a matter is in an individual's mailbox, the firm does not have the file. It has an arrangement with an employee, and that arrangement ends the day they resign, fall ill, take extended leave, or simply cannot be reached during the week the client needs an answer. Nothing about that is hypothetical. Every firm that has lost a lawyer mid-matter has spent the following fortnight exporting a mailbox and reading someone else's correspondence trying to work out what was promised to whom.

The exposure runs in three directions. There is the client-facing one, where a matter stalls because nobody can reconstruct where it had reached. There is the professional obligation one: rules on surrendering the client file when a retainer ends, and on what the file even consists of, vary considerably between jurisdictions, and firms in the United States, England and Wales, Canada and Australia should confirm what their own regulator expects rather than assuming a shared standard. And there is the security one, because a mailbox exported to a laptop so a colleague can search it is a copy of the file sitting outside every control the firm thought it had. A matter record that is genuinely firm-held, sitting in a system rather than a mailbox, removes all three at once.

Email Is Where Ethical Walls Leak

Firms that have built a wall around a matter tend to think in terms of the practice management system: this user cannot open that matter, and that is that. Email is the hole in that reasoning. A forwarded thread carries the substance of a walled matter to anyone at all, a reply-all pulls in a colleague who should never have seen it, and a search across a shared mailbox surfaces correspondence the wall was built to prevent. The wall stands and the information walks around it, which is worse than having no wall, because the firm believes it is protected.

This is why walls have to be enforced where the data is, not where the interface is. Casely enforces ethical walls at the server and data-access layer, so a walled user genuinely cannot reach a restricted matter by any path, including search, the calendar, or a link somebody forwarded them. Moving correspondence out of individual mailboxes and into the matter is what makes that enforcement real, because information sitting in the matter inherits the wall while information sitting in an inbox inherits nothing. The same logic applies to conflict checking: a conflict search that reads the full contact and matter history, across every role a party has played including on closed matters, only sees the parties it was told about, and parties introduced in an unfiled email are invisible to it.

What Filed Correspondence Does for Billing and Deadlines

The billing consequence is direct. A large amount of genuinely billable work happens in email, and email that never reaches the matter tends to become time that never reaches the bill. Correspondence captured against the matter as it happens gives you something to bill from that is not memory, which matters most for the short exchanges nobody records, the ones that add up to several hours a month per fee earner. When unbilled time is attached to the right matter, converting it into one itemised draft is a single step rather than an archaeology project, and firms working to insurer or corporate guidelines can export in LEDES 1998B without first reconstructing what happened from a thread.

The deadline consequence is less obvious and more dangerous. A very high proportion of date changes arrive by email, in a sentence buried mid-thread, agreed casually between counsel. If that message lives only in one inbox, the date it changed exists only there too. Attaching deadlines to the matter, with the next date tracked automatically, means the schedule and the correspondence that changed it sit in the same place, so the person covering next week is reading the same reality as the person who agreed to it. Combine that with a matter stage tracker configured to how the practice area actually runs and a colleague can see where a matter stands without opening a single message.

Making It Stick Without Turning Anyone Into a Filing Clerk

Policy does not change filing behaviour, defaults do. The realistic move is to make the low-effort path and the correct path the same path, which means composing from inside the matter, capturing inbound mail at first handling, and removing whole categories of email from the inbox entirely by moving updates, document delivery, and signatures into the portal. Where a rule is needed, keep it to one: correspondence that records advice, instruction, or a date does not live only in a mailbox. Rules that try to cover every case get read once and never applied.

Then measure the thing that tells you whether it is working, which is not how many emails were filed. It is whether a colleague can pick up any live matter and understand its current position without asking the responsible lawyer a single question. Run that test on a real matter, with a real colleague, and watch where they get stuck. The gaps will be specific and they will point at exactly which habit has not taken hold yet, which is far more useful than a compliance percentage that nobody trusts.

Where This Leaves the File

The reason email filing is worth solving properly is that it is not really about email. It is about whether the firm holds a complete record of its own work, or whether that record is distributed across individual mailboxes and reassembled under pressure whenever someone asks a hard question. Firms discover which of those is true at the worst possible moment, during a fee dispute, a handover, a regulator's enquiry, or the week after a departure, and by then the only available fix is expensive reconstruction. The cheap fix was available every day for years, at the moment each message was sent.

None of this requires anyone to become more organised. It requires the filing decision to move to the point where it costs nothing, which is composition, and it requires enough of the routine traffic to leave email entirely so that what remains is small enough to handle honestly. Do those two things and the matter file stops being an incomplete summary of the work and starts being the work itself, readable by whoever needs it, walled where it should be walled, and owned by the firm rather than by whoever happened to be on the thread.

If you want to see what that looks like in practice, the mechanics of holding correspondence, documents, deadlines, and stages in one matter record are covered in more detail on our matter management software page, and Casely is cloud-native with a free plan, so you can put a live matter through it before deciding anything.