CRM Data Hygiene for Law Firms: What to Track, What to Ignore

The standard failure mode of a law-firm CRM is not under-recording. It is over-recording. The firm rolls out forty-two fields per contact, partners fill in six of them, the data quality team chases the rest, and within a year the partnership stops trusting any of it.

The fix is not a stricter policy. It is a smaller schema.

The minimum viable record

For an intellectual-property firm, the contact record that actually drives business development is short:

• Name, current title, current organisation
• Role classification: in-house, private practice, foreign associate, inventor, intermediary
• Relationship owner inside the firm
• Last meaningful interaction (date, channel, owner)
• Source of relationship — how the firm came to know this person
• Known matters worked on together
• Open notes from the last conversation

That is the spine. A partner can look at that record and know what to do next. Anything beyond it has to earn its place.

The fields that produce friction without value

Some fields look useful and are not. They cost partner time at entry, age badly, and rarely change a decision. The recurring offenders:

• Lead score computed from arbitrary point values. Partners ignore it.
• Industry taxonomy with thirty options. Three would do.
• Free-text "interests" field. Nobody fills it; nobody reads it.
• Birthday and anniversary. Polite, never load-bearing.
• Preferred contact method. Stale within a quarter.
• Custom stages that mirror an internal process nobody outside BD recognises.

The test is simple. If the field would not be cited in a pitch decision, an origination credit dispute, or a conflicts check, it does not belong on the main record. Put it in a sub-object or do not capture it at all.

What actually predicts outcomes

The fields that predict instructions are unglamorous. Recency and depth of contact. Whether the relationship is held by one partner or several. Whether the contact has moved in the last twelve months. Whether the firm has been instructed by the contact's previous employer. Whether there is an open signal — a filing, a hire, a funding round — that creates a near-term reason to reach out.

These are the inputs to the BD function described in the four functions of legal BD. Most of them can be captured automatically if the system is wired to email, calendar, and the relevant registers. The partner should not be typing what the system can see.

The discipline

A clean CRM is not a CRM that is full. It is a CRM where every field has a reason to exist and a process that keeps it current. The exercise worth running once a year:

1. List every field on the contact and matter records.
2. For each, identify a decision it informs.
3. If no decision, retire the field.
4. For the survivors, identify how the field gets updated — automatic, partner entry, BD team entry — and whether that route actually works.
5. Anything updated by "everyone, in theory" gets reassigned or automated.

Firms that do this honestly cut their schema by half and see trust in the data rise immediately. The partner who opens a record and finds three accurate fields will return. The partner who opens a record and finds twenty fields, half of them stale, will not.

The relationship to adoption

Hygiene is not a separate problem from adoption. It is the same problem. Partners adopt systems they trust. They trust systems where the data they see is current and the data they enter is light. The shortest path to adoption is a smaller schema, captured mostly automatically, with the partner asked to confirm rather than create.

That is the operating principle. Everything else — dashboards, scoring, reporting — works only if the underlying record is something the partnership believes.

Related: The CRM as the Operating System of a Modern Law Firm. The Managing-Partner Adoption Playbook. Why Legal CRMs Die at the 90-Day Cliff.