Jump to section
The morning we lost a limitation date to a stale spreadsheet is the morning I stopped believing our system worked. It had never really worked. It had just failed without anyone noticing until it mattered.
I run operations at a six-lawyer firm in southern Ontario. Over four years we adopted three different ways of managing our practice, abandoned two of them, and limped through a migration that nearly took the office down with it. I am writing this because every vendor demo I sat through skipped the part I most needed to hear: what actually goes wrong, and which questions you only think to ask after you have been burned.
The Spreadsheet Years and Why They Ended
In the beginning there was a shared drive and a folder called MASTER. Inside MASTER lived a workbook with tabs for matters, key dates, trust balances, and a colour-coded thing our bookkeeper maintained that nobody else could read. It was free, it was flexible, and for a while it was enough.
The cracks were small at first. Two people editing the same row. A key date typed into the wrong client's line. A trust figure that matched the bank one week and drifted the next because someone forgot to record a disbursement. We patched each problem with a new rule and a stern email.
Then we missed a filing window on a civil matter. Nobody had touched the date, but the reminder lived in a cell with no alarm attached to it, and the file simply fell out of anyone's line of sight. We caught it in time to fix it, barely. That was the end of the spreadsheet years. Not because the tool was bad, but because it asked every human to be perfect every day, and humans in a busy office are not.
Warn. A spreadsheet stores information beautifully. It does nothing on its own. If your deadline system depends entirely on someone remembering to look, you do not have a system, you have a habit.
The First Platform Nobody Used
Our first real platform was chosen the way panic buys usually are: fast, and by the person most rattled by the near miss (me). I watched a polished demo, saw a calendar that pushed reminders, and signed up that afternoon. The problem was that I evaluated it alone.
The software was genuinely capable. It was also built for a workflow that did not match ours. Intake took eleven fields where our old habit took three. The billing module assumed a structure our bookkeeper did not use. Within a month the lawyers had quietly reverted to email and the shared drive, and I was the only person logging in.
A tool nobody opens is worse than no tool, because you are paying for it and still doing the work the old way. We let the subscription lapse and told ourselves we had learned our lesson. We had learned half of it.
A tool nobody opens is worse than no tool. You pay for it and still do the work the old way.
The Migration That Almost Derailed Us
The second real attempt was more careful. We involved the lawyers, ran a trial, checked that intake and billing fit how we actually worked. On selection, we did well. On the move itself, we nearly came apart.
Nobody warned us how much of a firm's history is trapped in its old system. We had years of matters, contacts, notes, and closed files. The export gave us a tangle of CSV files with cryptic headers, mismatched date formats, and trust entries that refused to reconcile against our records. For three weeks two of us worked evenings cleaning data, and we still went live with gaps we found for months afterward.
The lesson was not that migration is hard, though it is. The lesson was that the shape of your data on the way out matters as much as the features on the way in. If you cannot get a clean, complete export from a system, you do not fully control your own data. We now treat a workable export path, whether through a proper Data Import step or well-documented file formats, as a requirement rather than a nicety.
Tip. Before you commit, ask the vendor for a sample export of your own trial data, then try to read it. If the columns make no sense to you now, they will make no sense to whoever does your next migration.
The Questions We Wish We Had Asked First
By the third attempt we had a list. Not a feature checklist, those are easy and mostly beside the point, but a set of questions that surface how a tool will behave once real life gets messy.
| What we asked | Why it mattered |
|---|---|
| Who on staff has to change how they work, and by how much? | Adoption dies at the person with the steepest change. |
| Can we export everything, cleanly, on our own? | You will leave eventually. Plan the exit before the entrance. |
| Does trust accounting match law society rules for our province? | Generic bookkeeping is not the same as compliant trust handling. |
| What happens to a deadline when the responsible person is away? | Reminders must reach more than one set of eyes. |
| How does a new matter actually get created, start to finish? | Friction at intake means the system gets skipped at intake. |
The best question turned out to be the plainest: show me a lawyer opening a file, adding a note, and finding it again next week. If that loop is smooth, most of the rest follows. If it is clumsy, no dashboard will save you. For the wider debate on where software genuinely earns its keep, our take on the reality check every firm should run covers the same instinct: judge the tool by the boring daily task, not the glossy feature.
What Finally Made It Stick
The platform we settled on was not the flashiest of the three. What made it stick was fit and follow through. Intake matched how our clerks already worked, so nobody had to relearn their morning. Matters, time entries, and invoices lived in one place, which meant a lawyer could answer a client's billing question without three phone calls. And the trust side reconciled the way our bookkeeper needed it to, without a spreadsheet hiding underneath.
We also changed how we rolled it out. One partner championed it out loud. We moved active matters first and left the archive for later. We gave people two weeks where the old way and the new way ran side by side, then we turned the old way off, firmly, because a permanent fallback guarantees a permanent split. A1 CMS happened to be the tool that fit our shape, but the process would have mattered no matter which name went on the invoice. If you are weighing platforms now, our longer guide to evaluation and the plain pricing page are both honest places to start.
If I could hand my earlier self one page, it would say this: the software you can talk yourself into during a demo is rarely the one your firm still uses a year later. Choose for the ordinary Tuesday, not the sales pitch. Involve the people who will actually type into it. Test the way out before you commit to the way in. We lost a lot of evenings and one very bad morning learning that. You do not have to.
