CRM for a law firm: automating client work
A law firm runs on three resources: lawyers’ time, deadlines, and client trust. Each of them is easy to lose in Excel. Someone forgot that the appeal deadline expires on Tuesday. Someone worked 14 hours on a matter and billed 6, because they no longer remember the rest. Someone opened a case against a company the firm advised two years ago. A classic sales CRM does not work here: it is built around the deal, while legal business is built around the matter. Let’s look at what a CRM for a law firm should look like so that it actually closes these risks, instead of just neatly displaying a list of contacts.
Matter management: why a law firm CRM is built around the matter, not the deal
In a regular CRM the central entity is the deal: it has an amount, a pipeline stage, and a closing date. In a law firm the central entity is different — it is the matter. A matter can last two years, pass through three instances, involve ten hearings, fifty documents, two lawyers and a paralegal, and still have no “deal amount” until the very end. That is why matter management starts with the right record: case number, type (commercial dispute, insolvency, corporate support, family, criminal), client, opposing party, court and judge, instance, responsible partner, fee model, status. Everything else — time, documents, expenses, invoices, hearings — is attached to the matter, not to the contact.
A practical test that matter management is set up correctly: a partner opens a single record and sees everything within thirty seconds. How many hours have been logged, how many of them have been billed, how much has been paid, what the nearest deadline is, which document was filed last, what the court fees and expert costs are, which lawyer is running the matter and whether they are overloaded. If getting that answer requires opening three files, messaging an assistant and calling a lawyer — the firm does not have matter management, it has a folder with documents. The difference between those two states is measured in money: unbilled hours and missed deadlines cost more than any system.
Time tracking and billing: hourly, fixed fee, and everything in between
The main money leak in a law firm is unrecorded time. A lawyer reads the opponent’s documents, spends half an hour on the phone with the client, drafts a motion — and forgets to log any of it, because by evening there is no energy left. Over a month, 10 to 25 percent of billable time disappears, and that is a direct loss. So time tracking must be instant: the lawyer opens a matter, hits “start,” and the system counts; or later adds an entry in two clicks straight from a phone or a Telegram bot, picking the matter from their list of active ones. Every entry is a matter, a date, a duration, a description of the work, and a billable flag. The description matters: those very lines later form the invoice breakdown the client reads.
Next the system separates the two fee models. Hourly: hours are multiplied by the rate of the specific lawyer, because a partner and a junior cost differently, and at the end of the month an invoice with a day-by-day breakdown is assembled automatically. Fixed fee: the invoice does not depend on hours, but hours still have to be logged — otherwise you will never learn that corporate support at 30,000 hryvnia a month eats 40 hours and is in fact a loss. This is exactly where a CRM shows what the eye cannot see: it compares the matter budget with the time actually spent and warns you when the fixed fee is burning out. Add retainers and deposits: the system draws down the deposit and signals when it is nearly empty.
Deadlines and procedural terms: this is where legal automation starts
A missed procedural deadline is not an “inconvenience” — it is a lost case, a damaged reputation, and a potential claim from the client against the firm itself. That is why deadlines cannot live in someone’s head, in a notebook, or even in a shared calendar nobody opens. In a CRM a deadline is a separate entity with a matter, a type (filing an appeal, a response, a motion, complying with a ruling), a due date, and a reminder rule. The rule is simple and multi-level: a warning 14 days before, then 7, 3, 1 day before, and on the deadline day itself — with the last two going not only to the responsible lawyer but to the partner as well. The reminder arrives where the person actually lives: in Telegram, not in an inbox where it will drown among newsletters.
The second layer is hearings. Date and time, court, room, judge, who attends, which documents to bring. A reminder the evening before and two hours ahead. After the hearing the lawyer records the outcome with a single message to the bot: adjourned, granted, denied, next date — and the system immediately creates a new deadline. The third layer is oversight: every Monday morning the partner receives a digest with all the firm’s deadlines for the next 14 days and instantly sees who has too many. That is legal automation in its simplest and most useful form: not “AI writes your lawsuits,” but a system that will not let you forget what must not be forgotten.
Documents, pleading templates, and conflict-of-interest checks
A lawyer spends hours digging up last year’s statement of claim, copying it, and manually replacing names, dates and amounts — and it is at exactly this step that another client’s name ends up in the document. The right approach is a template library with fields. The system takes data from the matter record — the parties, company registration numbers, the court, the case number, the amount claimed — and fills them into the template of a claim, a contract, a power of attorney, or an act. The lawyer gets a draft in ten seconds and works on the legal position instead of on “insert here” placeholders. Every version of the document is stored next to the matter with a date, an author, and a status: draft, approved, filed.
A separate and badly underrated feature is the conflict-of-interest check. Before taking on a new client, the system searches the entire database by company name and registration number, and by the names of related individuals: has this entity ever been an opposing party, has the firm advised it before, is it connected to a current client. That is a five-second machine check instead of “I don’t think we ever dealt with them.” The result is stored as a record in the file: check performed, no conflicts found, date, who performed it. If a conflict exists, the record is locked until a partner decides. A single such check pays for the whole system.
Confidentiality, invoices, and analytics: a lawyer’s CRM that shows profit
Attorney-client privilege is not a slogan but a technical requirement. Access rights must be set at the matter level, not the system level: a lawyer sees only the matters they are assigned to, an assistant sees documents but not finances, a partner sees everything. Data is encrypted and access is logged: who opened what, and when. The question of where to store it is decided deliberately rather than by default: a private server or a rented server in a reliable jurisdiction, not a free cloud drive on an assistant’s personal email. Two-factor authentication is mandatory, a departing employee is cut off within a minute rather than “sometime later.” Backups run daily — and should be tested by an actual restore at least once.
The financial loop is closed by invoices and acts. At the end of the month the system assembles an invoice from logged hours and expenses, adds the breakdown, generates the completion act from a template, sends it to the client, and sets a reminder if payment has not arrived within seven days. Receivables stop being a mystery. On top of that sits the analytics everything was built for: profitability by matter type (it turns out insolvency work yields a 60 percent margin while family cases yield 15 for the same effort), workload and output per lawyer, the real effective hourly rate on fixed-fee contracts, the share of billable time. These are exactly the systems we build at Devlly — not a universal box, but a lawyer’s CRM shaped around a specific firm’s processes, with a Telegram bot for time tracking and reminders.