Client and matter data as the system of record

This lesson covers what data to trust, how to measure its quality, and the benchmarks that tell you if your firm's client and matter data is fit for purpose.

Why this data is structurally hard

Professional services firms (law firms, accounting and audit firms, consultancies, advisory boutiques) sell time and expertise, not units of inventory. Their core data objects are not products, they are relationships and engagements:

• Client: the legal entity or economic buyer relationship.
• Matter (legal) or engagement (consulting/accounting): a specific piece of work, with its own budget, team, and scope.
• Contact: an individual person tied to a client, often with multiple roles across matters.

Two structural features make this data hard to keep clean:

1. Client hierarchies are genuinely complex. A single corporate family (say, a private equity portfolio) can have a parent, subsidiaries, joint ventures, and individual executives who are also personal clients. If the CRMCRMCustomer Relationship Management: software and strategy to manage and analyse customer interactions throughout their lifecycle.Voir la définition complète → models "client" as a flat list rather than a hierarchy with parent-child relationships, cross-sell visibility collapses: the tax team doesn't know the M&A team is already engaged with the same ultimate parent.

2. Matter/engagement coding happens under time pressure. A new matter often gets opened same-day, by whoever is fastest, using free-text client name fields. Without enforced lookups against a master client list, this is where duplicates are born.

The core datasets that matter

| Dataset | System of record candidate | Key fields |

|---|---|---|

| Client master | CRMCRMCustomer Relationship Management: software and strategy to manage and analyse customer interactions throughout their lifecycle.Voir la définition complète → (e.g., Salesforce, Intapp) or a dedicated client data platform | Legal entity name, tax ID/EIN, industry (NAICS/SIC code), parent-child hierarchy, billing address |

| Matter/engagement master | Practice management system (e.g., Aderant, Elite 3E, SAP for consultancies) | Matter ID, client ID (foreign key), practice area, responsible partner, open/close date, conflict-check status |

| Contact/relationship data | CRMCRMCustomer Relationship Management: software and strategy to manage and analyse customer interactions throughout their lifecycle.Voir la définition complète → | Person ID, role, associated client(s), relationship owner |

| Time and billing | Practice management / ERP | Matter ID, timekeeper, hours, rate, write-offs |

| Conflicts data | Conflicts database (often bolted onto practice management) | All parties, adverse parties, related entities |

The critical design decision: which system owns the client entity, and which systems merely reference it. In mature firms, the CRMCRMCustomer Relationship Management: software and strategy to manage and analyse customer interactions throughout their lifecycle.Voir la définition complète → or a dedicated master data managementmaster data managementMaster Data Management (MDM) is the discipline of creating and maintaining a single, consistent, trusted version of an organization's core business entities like customers, products, and suppliers.Voir la définition complète → (MDMMDMMaster Data Management (MDM) is the discipline of creating and maintaining a single, consistent, trusted version of an organization's core business entities like customers, products, and suppliers.Voir la définition complète →) layer owns the client ID, and practice management, billing, and conflicts systems all pull from it via integration rather than allowing local free-text entry. Firms without this end up reconciling client lists in spreadsheets, a known failure mode.

For a practical reference on entity data structuring, the GLEIF Legal Entity Identifier (LEI) framework is a useful external standard: a 20-character code uniquely identifying legal entities globally, increasingly used by banks and large corporates. Some firms are adopting LEI as a cross-checked client identifier precisely because internal client IDs are unreliable.

Field-level standards that prevent duplicates

Duplicate client records are almost always traceable to a small number of ungoverned fields. Fix these four and duplication rates drop sharply:

• Legal entity name: enforce a canonical source (state of incorporation filings, or a data enrichment vendor like Dun & Bradstreet) rather than free text. Store "doing business as" (DBA) names as aliases, not as separate records.
• Tax identifier: US Employer Identification Number (EIN) or, for individuals, a masked internal ID; in Europe, VAT number or national company register ID (e.g., Germany's Handelsregister number, UK Companies House number). This is your best deduplication key because it's nearly unique and stable.
• Industry classification: use a standard taxonomy (NAICS in the US, NACE in the EU) rather than each practice group inventing its own labels. This is what makes cross-sell analytics possible: "show me all clients in NAICS 522110 (commercial banking) not yet using our regulatory practice."
• Parent-child relationship field: mandatory for any client with known corporate affiliates, populated at intake, not left blank "for now."

A minimal matching logic

Deduplication tools typically combine deterministic and fuzzy matching. A simplified pseudocode version:

This is the logic underneath MDMMDMMaster Data Management (MDM) is the discipline of creating and maintaining a single, consistent, trusted version of an organization's core business entities like customers, products, and suppliers.Voir la définition complète → tools like Reltio or Informatica MDMMDMMaster Data Management (MDM) is the discipline of creating and maintaining a single, consistent, trusted version of an organization's core business entities like customers, products, and suppliers.Voir la définition complète →, commonly deployed in large firms' client data cleanup projects.

Governance and quality metrics that matter here

Generic data-quality dimensions (completeness, accuracy, consistency, timeliness) need sector-specific translation:

• Duplicate rate: percentage of client records identified as duplicates during a periodic audit. Large firms running first-time MDMMDMMaster Data Management (MDM) is the discipline of creating and maintaining a single, consistent, trusted version of an organization's core business entities like customers, products, and suppliers.Voir la définition complète → cleanups commonly find duplicate rates in the 5 to 15% range on client master data (estimate, based on common industry cleanup project findings; actual rate is firm-specific).
• Hierarchy completeness: percentage of clients with a populated parent-child link where one is known to exist. Low completeness directly predicts missed cross-sell.
• Matter-to-client linkage integrity: percentage of matters with a valid, non-null foreign key to a client master record. This should be effectively 100%; anything less means matters are "orphaned" and invisible to relationship reporting.
• Conflict-check turnaround time: hours or days from new-matter intake request to cleared conflict check. This is a governance metric with direct risk implications under professional conduct rules (in the US, ABA Model Rule 1.7 on conflicts of interest; in the UK, SRA Standards and Regulations).
• Golden record adoption rate: percentage of downstream systems (billing, conflicts, CRMCRMCustomer Relationship Management: software and strategy to manage and analyse customer interactions throughout their lifecycle.Voir la définition complète →) actually consuming the master client ID rather than a local copy.

Analytics and benchmarks

Once the underlying data is trustworthy, a few analytics become possible that are otherwise fiction:

• Cross-sell penetration: number of practice areas or service lines used per client family, tracked against total available service lines. A firm might find its top 50 client families use an average of 2.3 of 12 available service lines, a concrete white-space number for business development, but only computable if hierarchy data is clean.
• Client concentration by ultimate parent: revenue rolled up to the top of the corporate hierarchy, not just the billing entity. Without hierarchy rollups, concentration risk is understated, a real issue when firms assess dependency on a single economic client.
• Matter profitability trend: realized rate versus standard rate over time, requires clean matter-to-client-to-timekeeper linkage.

A simple worked example: if a firm has 4,000 active client records and a data audit finds 420 are duplicates of another record in the system (a 10.5% duplicate rate, consistent with the commonly cited industry range for unremediated client masters), then any cross-sell report built before cleanup understates true relationship value for every affected client family, because revenue and matters are split across multiple "ghost" records instead of rolling up to one.

Key Takeaways

• Treat the client master as a governed system of record with one owning system; every other system (billing, conflicts, practice management) should reference it, not maintain its own copy.
• The highest-leverage fields for preventing duplicates are legal entity name, tax ID (EIN/VAT/company register number), industry code (NAICS/NACE), and parent-child hierarchy links.
• Track duplicate rate, hierarchy completeness, matter-to-client linkage integrity, and conflict-check turnaround as core governance metrics, not just abstract "data qualitydata qualityThe degree to which data is fit for purpose: accurate, complete, consistent, timely, valid and unique. Poor quality data undermines analytics, reporting and AI.Voir la définition complète →" scores.
• Cross-sell and concentration analytics are only as good as hierarchy rollups; a 10% duplicate rate (a realistic, commonly cited estimate for unremediated systems) can materially distort both.
• Conflict-of-interest checks are a data-quality issue with direct professional conduct consequences (ABA Rule 1.7, SRA Standards), not just an operational nuisance.