Automate conflicts at intake first, then time capture, then pre-bill review, then e-billing, then trust accounting, then docketing. That order is by payback, not by difficulty. Conflicts is first because every matter passes through it and it stands between a signed client and billable work. Docketing is last not because it matters least but because the firms that need it most already have a person whose whole job it is.
What happens in the video
- 0:00The six jobs, in payback order
- 0:38One: conflicts at intake
- 0:55Two: time capture
- 1:12Three: pre-bill review
- 1:31Four: e-billing
- 1:45Five: trust accounting
- 1:54Six: docketing
- 2:04Start with one
Systems on screen
ClioiManage WorkAderant ExpertMicrosoft TeamsMicrosoft OutlookPACERLawPayQuickBooks
Every firm does it differently. That is the point
The objection is real. Two firms of the same size in the same city run intake differently, bill differently, and reconcile trust differently. Any product that arrives with a fixed process is wrong on arrival, and every firm that has been sold one knows it.
But the variation is in how, not in which. There is no firm that has replaced conflicts checking with something else. There is no firm without a pre-bill. The list of jobs is the same everywhere; the way each one is done is local, and local is exactly what a recording captures and a template cannot.
So the question is not whether the back office can be automated. It is which of the six to start with, given that they are not equally expensive to leave alone.
The six, in payback order
1. Conflicts at intake
Five searches, four systems, before anyone opens a file. Every matter passes through it, every hour it sits is an hour of unstarted work, and the search itself is entirely mechanical up to the clearance decision.
Watch the whole check run across Clio and iManage, including the closed 2023 matter the CRM knew nothing about. The economics behind it are in new matter intake and conflicts at scale.
2. Time capture
Every email, document and call turned into a time entry the same day. Not because timekeepers are lazy, but because reconstructing Tuesday on Friday loses two tenths an hour, every time, at every firm that has ever measured it.
This is second because it is pure recovered revenue on work already done. See automated time capture for law firms and contemporaneous time entry.
3. Pre-bill review
Every entry checked against the client's own guidelines before the partner sees it. Block entries split from the record, vague narratives rewritten from what happened, wrong task codes corrected, and the entry nothing can source flagged rather than invented.
Watch a month of time reviewed against fourteen pages of outside counsel guidelines.
4. E-billing
LEDES files that do not come back. Most rejections are a code, a cap, or a missing field, and all three are checkable before the file leaves the firm rather than sixty days later on an appeal nobody bills for.
The full argument, and what the compliance tax actually costs in realization, is in outside counsel guidelines and e-billing. The code set itself is at the UTBMS code lookup.
5. Trust accounting
Three-way reconciliation every month, with the exceptions listed instead of hunted. The reconciliation is arithmetic across LawPay, the bank, and the general ledger. The judgment is what to do about an exception, and that stays with the person who signs.
6. Docketing
Court notices read, deadlines calculated under the right rule set, calendars updated, and the notice filed to the matter. Last in the order because most firms that carry real docketing risk have already put a person on it, which means the payback is quieter even though the downside is the largest on this list.
None of these was complicated
Look at the six together and the striking thing is how ordinary they are. None of them needs a model to reason about the law. They need somebody to do the job on screen once, saying what they are doing, and then for it to happen the same way every time.
They were never automated because they were never written down. They are the parts of a firm that live in the head of the person who has done them for eleven years, which is also why they break when that person is on holiday.
Start with the one that costs you evenings
The order above is a default, not a prescription. If your bottleneck is the pre-bill because a partner spends the last week of every month on it, start there. The point of the order is that these compound: clean time entries make pre-bill review cheap, and a clean pre-bill makes e-billing stop bouncing.
For the strategic version of this argument, without the screens, see workflow automation for law firms and AI and the business of law.
The rest of the series
Six films, each one a real run in the tools the work already lives in. Nothing sped past what you can read.
More for law firms
- How to run a conflict check automatically (2:23). Clio contacts and matters across every spelling, then full text search inside iManage documents, then the memo and the hold.
- How to automate pre-bill review (2:47). Forty-one entries read against fourteen pages of outside counsel guidelines. Three fixed, one flagged, nobody reading line by line.
The advisory firm track
- RIA back office automation: the six workflows (2:31). The six jobs that grow with the book, in the order they give hours back, with two of them shown end to end.
- How to auto-fill custodian forms from your CRM (2:39). Two forms, forty-two fields, eleven facts already in the CRM. One field left empty on purpose.
- How to turn a client meeting into CRM notes (2:21). A forty minute review becomes a note in the firm's format, four dated tasks, a logged recommendation, and a draft the advisor sends herself.
Start with one
Pick the job that costs you evenings
Show Caddi how your firm does it, once. It runs it the same way after that, in the systems you already have.
Frequently asked questions
What should a law firm automate first?
Conflicts at intake. It is high volume, every matter passes through it, it is entirely mechanical up to the clearance decision, and it is the thing standing between a signed client and billable work. After that: time capture, pre-bill review, e-billing, trust accounting, and docketing, in that order.
Why is legal automation usually framed as impossible?
Because firms say every firm does it differently, which is true and is also why nobody has tried. The variation is in how each firm does the job, not in which jobs there are. Every firm runs the same six, and the differences are exactly what a demonstration captures and a generic template cannot.
Is this the same as legal AI for drafting or research?
No. This is the business of law rather than the practice of it: intake, billing, compliance, and the operations around the work. It does not touch legal judgment, advice, or drafting, which is why it is the part of a firm that can be automated without a partner having to supervise the output line by line.
How long does it take to automate one of these?
The setup is a recording of the work being done once, narrated, usually fifteen minutes to an hour depending on the job. The films in this series show the whole loop from that recording to a run, which is the honest measure of what is involved.
Do we need to replace Clio, iManage or Aderant?
No. Every run in this series works inside the tools the firm already has, because the work already lives there. The agent is shown the job in your screens and does it in your screens.