Drafting Engagement Letters and Standard Documents From the Matter Record

An associate at a small real estate and estate planning firm needs an engagement letter and a standard document request for a new client. She opens the matter that looks most similar, copies the letter, and runs find and replace on the names. The fee paragraph is from last year's rates. In paragraph seven, the previous client's first name survives because it was spelled with a different capitalization. A partner catches it in review, which is the good outcome, and on a busier week it could easily have gone out.

Drafting from the last similar file is how most small firms produce routine documents, and it carries two problems. It takes lawyer and paralegal time that rarely adds value, and it creates a steady low-level risk of carrying one client's details into another client's document. Both can be fixed with a disciplined template library and a matter record clean enough to fill it.

This article covers which documents suit automation, what needs fixing in your templates and data first, and what the drafting flow looks like inside Clio, MyCase or a similar practice management system.

Which documents are good candidates

The best candidates are documents that are produced often, follow a predictable structure, and vary mainly in facts that already exist in the matter record:

• Engagement and non-engagement letters
• Standard correspondence: acknowledgments, document requests, status letters, closing letters
• Routine transactional documents, such as simple leases, NDAs and standard closing documents
• Estate planning documents with well-defined variations, drafted under careful review
• Court forms and standard pleadings where the structure is fixed and the facts change

Bespoke advice letters, complex agreements and anything that turns on judgment remain drafting work for lawyers, though even there a first draft of the boilerplate sections can save time.

Fix the templates and the data first

One approved template per document type

Gather the versions of each document currently in use and agree one approved template, owned by a named lawyer. Store templates in one place with version control, and stop the habit of drafting from old matters. If people keep using old files because the template is out of date, that's a sign the template owner needs to keep it current.

Mark variables and conditional clauses explicitly

Every piece of text that changes between matters should be a marked field: names, addresses, dates, fees, property details, roles. Clauses that only apply in some situations, such as a clause for clients with minor children or a provision for a joint representation, should be marked with the condition that triggers them. This is the work that makes automation reliable, and it often reveals inconsistencies between versions that nobody had noticed.

Clean up the matter record

Templates fill from the matter record, so the record needs the right fields, filled consistently. In Clio and similar systems, custom fields can hold practice-specific facts such as property address, closing date or number of beneficiaries. Agree which fields are required at matter opening, and fill them from intake instead of retyping.

Agree naming and storage

Decide how generated documents are named and where they're saved in the matter, so anyone can find the current version and see what was sent.

What the drafting flow does

1. Choose the document from the matter. From within the matter, the lawyer or paralegal picks the document type. The right template is selected automatically based on the practice area and matter facts.
2. Fill the facts. Fields are filled from the matter record and intake answers: parties, addresses, dates, fees, property details. Conditional clauses are included or left out according to the matter's facts.
3. Draft the variable prose. Where a document needs a short narrative, such as a description of scope in an engagement letter or a summary of facts in a demand letter, an AI model drafts it from the lawyer's notes and the intake answers. Those sections are clearly highlighted for review.
4. Check before review. The system checks for empty fields, names that don't match the parties on the matter, dates that conflict, and any text that matches another client's details. Issues are listed at the top of the draft.
5. Lawyer reviews and edits. The lawyer reviews the draft in Word, edits as needed and approves it. Nothing goes to a client without that step.
6. File and send. The approved document is saved to the matter with the agreed naming, and sent for e-signature where needed, with the signed copy filed automatically on return.

Clio's built-in document automation and dedicated tools like HotDocs already handle field merging well, and if you have one of them, it's the natural base to build on. The additions that make the bigger difference are the conditional logic, the drafted narrative sections, the pre-review checks and the link back to intake. Our one-click paperwork system adds those around the tools you have. The law firms page shows how drafting fits alongside intake, client updates and the precedent assistant.

Professional responsibilities still apply

Using AI in drafting doesn't change a lawyer's obligations. The ABA's Formal Opinion 512, issued in July 2024, addresses how existing duties apply to generative AI tools, including competence, confidentiality, communication with clients, supervision and reasonable fees. In practical terms, that means understanding what the tool does, protecting client information, reviewing its output properly, and billing honestly for the time the work actually took. Your own jurisdiction may have issued further guidance, and it's worth reading before you roll anything out.

A worked example

Consider the real estate and estate planning firm from the opening. The team gathers the engagement letters currently in use and finds several versions with different fee wording. A partner agrees one approved template for each practice area, marks every variable, and adds a conditional clause for joint representation of spouses. Matter opening now requires the fields those templates need, filled from the intake questionnaire. When the associate opens a new estate planning matter for a married couple, she selects the engagement letter, and the draft arrives with both clients' names and addresses, the current fee schedule and the joint representation clause included. The pre-review check notes that one spouse's middle name differs between the intake form and the matter contact, and she confirms the right one with the client. The partner's review is quick, because the only part that needs real attention is the scope paragraph, drafted from the associate's notes and highlighted.

What to watch out for

• Client confidentiality. Know where any drafting tool processes and stores client data, whether it's used to train models, and who at the vendor can access it. Choose tools that keep data within your control.
• Template drift. Templates go stale as rates, regulations and preferences change. Give each template an owner and a review date.
• Over-trusting the draft. Clean formatting makes errors harder to spot. Review generated documents as carefully as anything drafted by a junior, especially the narrative sections.
• Execution formalities. For wills, deeds and other documents with signing and witnessing requirements, make sure the template and signing process follow the formalities in your jurisdiction, and don't route them through e-signature where it isn't valid.
• Data quality. If matter fields are filled inconsistently, generated documents will be wrong in ways that are easy to miss. Required fields and intake that writes straight to the matter record prevent most of this.

How to tell whether it's working

Track the time from request to approved document for your most common document types, before and after. Count the issues the pre-review checks catch, which shows both the value of the checks and where data entry needs tightening. Note how much reviewers change the drafted narrative sections, since heavy editing means the notes or instructions need work. And keep an eye on template usage: if people still draft from old files, ask why, because it usually points to a template that's missing a variation they need.

Where to start this week

List the ten documents your firm produces most often and, for each, note how it's produced today: from a template, from an old file, or from scratch. Pick the three that are produced most often from old files and build approved templates for them, marking every variable field and conditional clause. Then check that the matter record has fields for every variable those three templates need.

That work pays off even before any automation, because everyone drafts from the same approved version. Once the templates and fields are in place, generating first drafts from the matter becomes a small step, and review time goes into the substance of the document instead of hunting for the last client's name.

For a quick read on where drafting sits among your firm's other administrative work, run the free PATH report. Or book a call and we'll look at your templates with you.