To open a client's file at the firm, you first had to know the secret path. You navigated to the network drive — which was really the founding attorney's own desktop PC moonlighting as a server — and clicked down through three separate folders, each named Immigration Forms, nested one inside the other like Russian dolls. Buried at the center was a folder called Files - Open where the client documents lived. But the information you needed might not be there at all. It might live under Case Management, or in a caseworker's private folder, or in one of a dozen Google Sheets, each color-coded by a logic known only to its author, each listing the same client under a slightly different spelling of their name.
Matthew Knight found that out his first week on the job. Hired as a brand-new legal case manager at an 8-person immigration law firm, he was still being walked through how to find things when the scale of the problem hit him. "I almost had a panic attack," he says. "It dawned on me that there was no single place anything actually lived."
This was the firm's operating reality: a decade-plus of casework, spread across a system nobody had really designed, just allowed to accumulate. Abacus, the practice-management software at the center of it, had been sunsetting for years. The firm ran an old local copy that only tracked contacts and calendar entries. There weren't even enough Abacus licenses to go around: five seats for eight people, so someone would quietly log you out mid-appointment to grab a seat, and you'd come back to your desk locked out of your own case.
Every action on a case had to be logged in at least three places: a note in Abacus, a line on the printed progress sheet stapled to the paper file, and a row in the relevant spreadsheet. Inevitably, the different trackers drifted out of sync, and staff often had to check all of them to know what was really happening with their cases. And then there was the communication logjam. Client texting ran through a single shared iPhone. Fifty times a day, the same question rang out across the office: "Who's got the phone?"
The staff had stopped noticing. "They'd been doing it for years. It was just normal to them," Knight says. "Coming in from outside, it was almost unbelievable."
When the system failed
The dysfunction wasn't just inconvenient. This is immigration law, which means the stakes are often enormously high. "Nobody dies if we get a form wrong," Knight says. "But a rejected packet can be the difference between a client finishing their hearing and a family getting separated."
Which is exactly what the old system nearly caused. On some of his first cases, Knight drafted petitions using USCIS forms a colleague had saved into his folder — the standard onboarding ritual. But the forms were years out of date. Nobody caught it: not Knight, not the case manager training him, not even the attorney.
USCIS did. Ten cases bounced back, each requiring the whole packet to be redrafted on the current form and resubmitted. It was pure rework, at a firm whose quarterly filings — and therefore revenue — had already slipped from 99 the year before to 75. "It wasn't anyone's fault," Knight says. "The system was built so that you couldn't tell which version of a form was the right one. Search a template and you'd get hundreds of hits — live client copies, deprecated versions, someone's personal library — with no way to know which was current at a glance."
For Grace LaRocca, the firm's founding attorney and sole owner, that was intolerable. Her goal is simple: to run the best immigration firm in the world. She knows her clients come to her precisely because they can't navigate the U.S. immigration system alone, so her systems have to be the best.
An outsider with nothing to lose
When Grace asked her newest hire what he'd change at the firm, Knight told her the truth without hesitation. Reorganize the network drive into a real knowledge base. Clean up the duplicates and deprecated files. Kill the paper trackers. Get to a single source of truth.
"I think we all knew something was wrong, but none of us knew how to fix it. We needed someone to come in from the outside and see it with fresh eyes."
LaRocca often asked for suggestions, but her staff, heads-down under heavy caseloads, rarely had the bandwidth to give it. Knight, the new guy on the team with an MBA in progress, a Lean Six Sigma certification, and a history in tech implementation and process improvement, did. "I figured she'd respect the honesty," he said, laughing. "And if she didn't, I didn't have much capital to lose yet anyway."
She wasn't shopping for new software yet, but she was hungry to do right by her clients, and she gave the green light. Knight dove in to the project with enthusiasm. It wasn't long before he realized the whole system would have to be rebuilt, beginning with the aging software at the center of it all. He evaluated three potential platforms and built a three-year NPV/ROI model to compare them on total value, rather than just sticker price. Clio won on both value and adoption risk. The software was easier to use than its competitors, and the promise of dedicated migration support helped smooth the prospect of a difficult transition.
The migration, and the resistance
The cutover itself was disciplined. A Clio migration specialist moved the Abacus data over in about a week; Knight set a hard stop — nothing enters Abacus after 4 p.m. on cutover day — and stood up a bare-bones Google Sheet to catch updates in the gap, then folded them into Clio once it was live. He standardized the messy spreadsheets (with the help of custom AI workflows) into two clean master sets for upload. He gathered the firm's scattered templates and affidavits and converted each of them into Clio Draft documents with merge fields.
It was done in two weeks. No data was lost. The people were the harder migration.
Once Clio went live, staff members started to complain. The new client intake workflow felt long. In reality, it asked all the same questions as the firm's old intake form, but it required caseworkers to scroll down a single column in the middle of a screen instead of flipping over a familiar double-sided sheet of paper. One caseworker, Marisol, simply decided to skip it, running quick intakes and booking consultations without gathering all that information. A week later, clients started showing up for those consults with no file prepared.
Rather than mandate compliance, Knight sat with Marisol one-on-one and walked through the whole intake workflow end to end until she could see how it worked. More importantly, she saw the value. Clio used the information from that intake to draft the forms that would later be used for the client's immigration petition. "The resistance was really about uncertainty," he says. "The first impression was 'this is so long and unwieldy,' so people reached for the familiar. Once they see it work, that goes away."
What made it go away fastest, though, was texting. Clio's built-in text messaging let anyone pull up a client and message them from their computer. No setup, no shared iPhone, no more "who's got the phone?" — a real quality-of-life win the whole office felt on day one. From there, the wins compounded. One-click drafting of an affidavit or a background-check letter, always from the current template. Every email, text message, 3 a.m. client photo, note, and form living on the matter, in one place. A single source of truth.
Knight's favorite signal came later, and it wasn't a metric. "Caseworkers started finding things in the software I didn't know about and teaching me — a scheduling trick, a faster way to see whether a consult was booked. That's when I knew we weren't just at 100% adoption. We had buy-in. The implementation was a success."
The results
The firm is kept anonymous; all names except the author's are pseudonyms.