CCCasestack Cloud
Guides › Legal Operations

Switching Law Firm CRMs in Ten Steps Without Losing a Referral Source

Most firms switch CRMs without noticing they've lost referral credit, partner ownership and opt-out history along the way. This ten-step cut-over plan protects all three, with test scripts to run before go-live.

Casestack Cloud editorial team · November 2026 · 8 min read

What Breaks First in a Law Firm CRM Migration

The contacts usually make it across. The context around them often doesn't. In a typical law firm CRM migration, four things go missing: who referred the matter, which partner owns the relationship, what the contact agreed to receive (and opted out of), and where each open inquiry stood in the intake pipeline. You won't see the damage on day one. It surfaces a quarter later, when a referral report comes back empty, a partner finds a junior colleague emailing their client, or someone who unsubscribed years ago gets your newsletter again.

This plan has ten steps, and each one protects one of those four assets. It works whether you're moving from a legacy system such as LexisNexis InterAction to Intapp DealCloud, from spreadsheets to Clio Grow or Lawmatics, or from a general-purpose CRM like HubSpot or Salesforce to a legal-specific tool.

Advertisement

Step 1: Freeze the schema and take an inventory. Two to four weeks before export, stop anyone creating new custom fields, tags or picklist values in the old system. Then export a complete field list with fill rates. A field populated on a tiny fraction of records is a candidate to drop. A field on most records that nobody can explain needs an owner before you map it.

Mapping Referral Sources and Relationship Partners Field by Field

Step 2: Model referral source tracking as a relationship, not a text field. Older systems often store the referrer as free text ("J. Smith – Barclays"), a tag or a dropdown value. Most modern legal CRMs let a referral source be a linked contact or organization record, and that's what makes reports like "matters and fees by referrer" possible. Convert free text to linked records before import. Afterward it takes far longer.

Step 3: Map relationship-partner ownership explicitly. Many old systems record ownership implicitly, by who created the record or whose Outlook it synced from. The new system will assign whoever runs the import unless you say otherwise. Decide whether you need one owner or several (lead partner, secondary partner, business-development contact), and give each its own field.

AssetCommon legacy formatTarget formatRisk if mapped carelessly
Referral sourceFree text or tagLinked contact/organization recordReferral reports show zero; referrers go unthanked
Referral date and matter linkOften missingDate field plus matter or opportunity IDCan't attribute fees to referrers
Relationship partnerRecord creator or synced mailboxNamed owner field(s)Every record owned by the import user
Consent and opt-outsCheckbox, or held in the email platformPer-channel status with date and sourceRegulatory exposure; suppressed contacts mailed again
Intake stageCustom status valuesPipeline stage mapped one-to-oneOpen inquiries stall or vanish

Contact Data Cleanup Before Export, Not After

Step 4: Deduplicate and standardize in the source system. Cleaning before the move is cheaper, and it lets the people who understand the conflicts resolve them. In practice that means merging duplicate people across partners' synced address books, standardizing organization names ("HSBC", "HSBC Bank plc", "HSBC UK"), and flagging departed contacts rather than deleting them.

One rule matters more than the rest: when duplicates merge, keep the most restrictive consent status and every referral link. Merge tools often default to "newest record wins," which can quietly overwrite an opt-out or detach a referral history.

Hypothetical example: say a 60-lawyer firm exports 48,000 contacts and finds roughly a fifth are duplicates. Partner-by-partner review can easily take a few hours of each partner's time. Budget for it explicitly, or it won't happen.

Step 5: Export consent as evidence, not as a yes/no flag. Under GDPR-style regimes you may need to show when and how consent was obtained, or which lawful basis you rely on. Each consent record should carry the channel, status, date captured, source (web form, event, matter engagement) and, where you have it, the wording the person agreed to.

Rules differ by region, so check where your contacts actually sit:

  • EU and UK: GDPR and UK GDPR, plus ePrivacy rules (PECR in the UK) for electronic marketing. Legitimate interests may cover some business-to-business contact, but opt-outs must always be honored.
  • Canada: CASL generally requires express or implied consent for commercial email, and implied consent can expire.
  • Brazil, South Africa, Singapore, India: LGPD, POPIA, the PDPA (with its Do Not Call rules) and India's DPDP Act each set their own consent and notice requirements.
  • Australia: The Spam Act requires consent and a working unsubscribe facility.

Step 6: Merge suppression lists from every system. Opt-outs often live in Mailchimp, HubSpot or an event platform rather than the CRM. Combine them into one suppression list and import it first, so no campaign can go out before it's in place. If your contacts span jurisdictions, have data-protection counsel confirm which lawful basis applies to each segment.

Moving Open Intake Pipelines Without Dropping Live Inquiries

Step 7: Choose a cut-over method for in-flight intake. Inquiries that are mid-conflict-check, waiting on an engagement letter or booked for a consultation are your most time-sensitive records. Two approaches work:

  • Hard cut-over: Migrate every open opportunity on a Friday, stages mapped one-to-one, and route all web forms and intake phone notes to the new system from Monday. This suits firms with a modest number of open inquiries, say a hundred or so.
  • Parallel run: New inquiries go into the new system while existing ones finish in the old one over two to four weeks. It's safer for high-volume practices such as personal injury and immigration, but someone must own the old queue daily until it's empty.

Either way, carry over the referral source, assigned intake owner, next-action date and any conflict-check reference for every open record. Re-point web forms, chat widgets and call-tracking numbers the same day, and submit a test inquiry through each.

Test Scripts to Run Before Go-Live

Step 8: Run a full trial import into a sandbox. Use the complete dataset, not a sample, because edge cases only appear at volume. Step 9: Have named testers run scripted checks and sign them off.

TestHow to run itPass criterion
Referral reconciliationCompare count of matters by referrer, old vs. new, for the last three yearsCounts match; every variance explained
Top-referrer spot checkOpen your top 25 referrers and check their linked mattersAll links present, with correct dates
Partner ownershipEach partner reviews 20 of their key clientsCorrect owner; no records assigned to the import user
Opt-out suppressionPick 50 known opt-outs and add them to a test campaignAll 50 excluded automatically
Consent evidenceOpen 20 consented contactsDate, source and channel all populated
Intake routingSubmit a test inquiry through every form and channelLands in the right stage, with the right owner, within minutes
Merge integrityReview 30 merged duplicatesMost restrictive consent kept; all referral links kept

A failed opt-out test is a hard stop. Weigh every other failure against the launch date.

Cut-Over Day, Costs and the First 30 Days

Step 10: Go live, then keep the old system read-only for about 90 days. Don't cancel the old licence on cut-over day. You'll need it to settle disputes like "this referral was mine." Rerun the referral reconciliation at 30 days, once new matters are flowing through.

On cost, typical 2026 market ranges vary widely by contract and region. Small-firm intake tools such as Clio Grow or Lawmatics commonly run from tens of dollars to a couple of hundred dollars per user per month. Enterprise platforms such as DealCloud or InterAction are usually quoted per firm and cost considerably more. Migration services from a vendor or partner often fall in the low-to-mid five figures in USD for small and mid-size firms, and enterprise projects with heavy cleanup and integrations can cost well into six figures. Local pricing and tax differ, so confirm current figures with vendors.

Red flags in a vendor's migration proposal: no sandbox trial import, "we import your CSV" with no mapping workshop, no plan for consent history, and a fixed price that excludes deduplication.

Set the go-live date by one rule: if the sandbox can't pass the referral reconciliation and opt-out suppression tests, you're not ready. Move the date, not the standard.

Common questions

How long does a law firm CRM migration usually take?

For a small or mid-size firm, plan on roughly two to three months from freezing the schema to going live, though it varies. A large share of that goes on cleanup and partner review, not the technical import. Enterprise migrations with several integrations often take considerably longer, commonly several months.

Should we clean contact data before or after migrating?

Before. The people who can settle duplicates and ownership disputes know the old system, and merge rules are easier to control at the source. Cleaning afterward often locks in the very errors you were trying to remove.

Can we just re-ask all contacts for marketing consent after the move?

In many jurisdictions, emailing people to ask for consent is itself treated as marketing. Under UK and EU rules that can be a breach if they've already opted out. Migrate your existing consent and suppression records properly, and get advice from data-protection counsel before running any re-permission campaign.

What if our old system never tracked referral sources properly?

Rebuild what you can for the last few years from matter-opening forms, intake notes and partner interviews, and import it as linked records. Mark reconstructed entries as such. From go-live, make referral source a required field at intake.

Do we need the vendor's migration service or can we do it in-house?

In-house can work if you have a legal operations or IT lead with data-mapping skills and a contact database of modest size. With large databases, complex consent history or several integrations, a paid migration service usually pays for itself in avoided rework. Either way, insist on a sandbox trial import.

Who should sign off before go-live?

At minimum: the head of business development or marketing (referral and campaign tests), a partner representative (ownership tests), whoever is responsible for data protection in the firm (consent and suppression tests), and the intake lead (pipeline routing). Record each sign-off against the specific test it covers.

Written by the Casestack Cloud editorial team. Prices, features and rules change often - confirm details with the vendor or a licensed professional before you decide.