The failure people plan for in a software rollout is resistance: staff who prefer the old way, training that does not stick, features nobody uses.
The failure that actually causes damage is the middle period. For a fortnight or a month, some client information is in the old system and some is in the new one, and no single place holds a complete client file. The Code of Conduct requires that file to exist and to contain specified things. During a badly run transition, for some matters, it does not.
A thirty-day plan is really a plan for keeping that window short and bounded.
Days 1–7: configure against obligations, not preferences
The instinct is to configure the system to match how the practice currently works. Some of that is right, and some of it is reproducing habits that exist because the old software forced them.
Configure the compliance-bearing parts first, because they are what the system is for:
- Matter and client structure, including the ability to hold several clients against one matter — sponsors and applicants are separate clients, and one file may cover a family unit unless a member asks for their own.
- Document categories that map to the file's required contents, including places for records of third-party calls and evidence of returned originals.
- Deadline types with consequence classes.
- Client money, tracked per client and per purpose.
- Templates for the service agreement and its variations.
- Permissions, so access follows role and matter rather than defaulting to everyone.
Leave preferences — dashboards, colour coding, personal views — until the second week. They generate the most opinions and the least value.
Decide one thing now, in writing: the date from which the new system is authoritative. Everything else in the plan depends on it.
Days 8–14: pilot on matters that can absorb a mistake
Pick five to eight matters and one or two people. The selection criteria matter more than the number.
Choose matters with no deadline in the next month. The pilot will be slower than business as usual, and a hard date inside it converts a learning exercise into a risk.
Include one of each shape. A straightforward matter, a sponsored one with two clients, a family unit, and one with a difficult history. A pilot of five easy matters teaches you nothing about the fourth week.
Choose people who will say what is wrong. Not the most enthusiastic — the most candid.
Run those matters entirely in the new system, including notes, correspondence and billing. Half-running them proves nothing.
What you are testing is not whether people like it. It is whether a complete client file can be produced from the new system for each pilot matter. Take one and check it against the Code's list, item by item. Anything you cannot find is a configuration gap, and finding it now costs an afternoon.
Fix the workflow, not just the software
The pilot will surface things that are awkward. Distinguish two kinds.
Awkward because it is unfamiliar — leave it. Two weeks of use resolves most of it.
Awkward because the process was never designed — fix it now. Rollouts are the rare moment when a practice will accept a change to how something is done, and that window closes quickly.
Days 15–21: train on the work, not the features
Feature training produces people who know where the buttons are and still do the work in the old way.
Train on the tasks the practice actually performs, in the order they occur: open a matter, send the agreement, request documents, record a call, lodge, invoice, close. Have each person do it themselves rather than watch.
Three additions worth the time:
- A session on what changes about the record. Notes are contemporaneous now. Correspondence goes on the matter. Individual logins, always. This is the part that carries obligations, and OMARA's guidance on practice management duties lists regular training on Code requirements among the supervisory arrangements a practice should have.
- A session for unregistered staff specifically, covering what the system lets them do and where the boundary sits. New software does not extend what they may do.
- A written one-page reference for the five tasks people do daily.
Section 24 makes an agent responsible for the quality, accuracy and timeliness of work performed by others assisting them, and a transition is exactly when that quality dips. Plan for closer review during it.
Days 22–30: cut over, then hold the line
Move all active matters on the authoritative date. Do it in one movement rather than gradually — a gradual move is precisely the split-record state to avoid.
Make the old system read-only the same day. Not "we'll stop using it" — actually read-only, so nobody can update a matter there.
Route incoming correspondence to the new system from the same date, including anything arriving in personal mailboxes.
Run a daily check for the first week: any matter touched in the old system, any correspondence that did not reach a matter, any deadline not carried over.
Then hold. The pull back to old habits is strongest in week five, once the novelty and the support have both faded.
What to measure
Adoption is measurable and worth a weekly look for the first two months:
- What proportion of matters have activity in the new system this week?
- Are notes being recorded at all, and at the time?
- Is correspondence landing on matters, or still in mailboxes?
- Are invoices being raised in the new system?
- Where does anyone still use the old system, and why?
That last one is diagnostic. A persistent fallback almost always points at one task the new system does badly, and fixing that single task usually resolves the whole pattern.
The things that make rollouts fail
Four, in order of frequency:
No authoritative date. Without one, both systems stay half-alive indefinitely and nobody is wrong.
Configuring for everyone at once. The pilot exists so the configuration is wrong for eight matters rather than for the whole caseload.
Training before the system is configured. People learn a system that then changes, and lose confidence in both.
No owner. A rollout with no named person responsible reverts. It does not need to be the principal; it needs to be someone with the authority to decide.
Our guide to growing from solo to a team covers the point at which practices usually attempt this, which is often the busiest possible moment.
The rollout test
At day 30:
- Can you produce a complete client file from the new system for any active matter?
- Is the old system genuinely read-only?
- Is all correspondence reaching matters?
- Are notes recorded at the time, or in batches?
- Does anyone still have a reason to open the old system?
- Has anything been lost — a deadline, a note history, an attachment?
Where a system helps
The honest note to end on is that the rollout is not the hard part. Choosing well is.
A system that models what a migration practice actually does — several clients on one matter, agreements with versions, client money per purpose, deadlines with consequences, correspondence on the file — makes the rollout mostly configuration. A system that does not will make every week of the thirty an exercise in working around it, and the workarounds become the practice.
LodgeHQ is built around those structures. Our comparison of case management software for migration agents covers how to evaluate the alternatives; start a free trial and run a pilot matter through before you commit to anything.
Verify before you rely on it
Thirty days is a shape rather than a rule, and a practice with a large caseload or complex integrations should plan for longer. What should not stretch is the period during which two systems are both live.
This is general information for migration practices, not legal or IT advice. Where a rollout coincides with matters carrying hard deadlines, plan those matters individually rather than relying on the general sequence.