Marketing Attribution for PI Firms: Follow the Case, Not the Click

The two-dashboard problem

A 2026 episode of Lunch Hour Legal Marketing offers a useful warning. In A Tale of Two Dashboards, Conrad Saam and Gyi Tsakalakis described an $11,500 campaign that generated 64 leads. After reviewing the calls, the agency believed 27 were qualified. The firm said the campaign produced no signed cases.

The lesson is not that one side was right. The agency measured cost per lead. The owner cared about cost per client. There was no shared signal for what the firm considered wanted, what happened after intake, or how those outcomes returned to the marketing view.

When those definitions and records are disconnected, good marketing can be cut because intake failed. Weak marketing can survive because it generates activity. And the firm can change agencies without repairing the operating problem that follows it.

Aim for useful truth, not perfect certainty

A PI prospect may see a billboard, hear the firm's name from a friend, read reviews, search on Google, visit a location page, and call days later. Assigning all credit to the first or last recorded touch is tidy, but it is not the same as understanding the decision.

In his episode on why law-firm marketing ROI is difficult to track, Chris Dreyer recommends combining system evidence with an open-ended intake question: how did you hear about us? The structured trail shows what the system observed. The prospect's answer adds what they remembered. Neither should erase the other.

Attribution becomes more reliable at volume. One disputed case should not move a budget. Repeated patterns across source evidence, prospect recall, wanted-case rate, signed cases, and fees can.

The source should survive a changed phone number

In a Tip the Scales conversation, Maria Monroy raised a common failure. A prospect may call from an office phone or a relative's device, then give the intake specialist a different number for future contact. If attribution depends on matching one phone field later, the trail breaks.

Yani Smith explained that her team's Lead Docket setup preserves multiple data points: the referral and current URLs, UTM values, the page involved, the number used, and the contact information gathered by intake. The full conversation is available in her intake optimization episode.

The operating principle is simple: source evidence belongs to the opportunity, not to one mutable identity field. Update the preferred callback number. Keep the original trail.

How to use Lead Docket for attribution

Lead Docket can serve as the intake system of record, but installing it does not make attribution accurate. The configuration and the team's data discipline do that.

The Lead Docket mechanics behind the setup

Lead Docket's official glossary separates Marketing Source, Contact Source, and Opportunity Source. Its integration documentation supports creating Opportunities from sources such as website forms, web chat, answering services, lead providers, and social forms, with incoming values mapped to standard or custom fields.

For digital attribution, Lead Docket's Source Rules can assign Marketing Source and Contact Source from matching URLs or UTM values. Test overlapping rules carefully: the documentation says the most recently added matching rule takes precedence, and edits affect future records rather than rewriting historical ones.

When configuring an external source, enable enough raw detail to audit the mapping. Lead Docket's guide to integrating Opportunity sources explains how incoming fields can be shown in Opportunity notes and mapped into existing or custom fields. Send a test from every real path, including mobile and after hours, before trusting the reports.

What the owner should review

A useful attribution dashboard does not stop at leads, and it does not hide counts behind percentages. Review each source across the same operating and economic measures.

This is also where the idea of kept-case acquisition cost becomes useful. In a podcast discussion of PI firm unit economics, Anthony Lopez emphasized that acquisition cost, case fallout, time on desk, and the labor required to resolve a case all affect profitability. A signed case is an important milestone, not the final economic answer.

Lead Docket provides an Agency Report with leads, wanted leads, signed leads, and source breakdowns without financial information. Internally, the firm can use documented revenue-by-source and fee reports. Those reports become decision-grade only when source, status, and financial data are complete.

AI should reconcile the evidence, not invent it

Attribution creates a useful role for AI. An agent can compare source payloads, UTM values, referral URLs, call metadata, contact updates, intake notes, statuses, and downstream outcomes. It can flag missing fields, likely duplicates, conflicting classifications, and records that stopped moving.

It should not silently guess a source. The system should preserve the supporting evidence, state its confidence, and route uncertain matches for human review. That same boundary applies to the broader PI intake operating system: AI can detect, structure, route, remind, and measure. People own judgment, case acceptance, and the relationship.

Start with a discrepancy audit. Compare the last 50 inquiries across the original source evidence, Lead Docket record, signed matter, and available fee data. The repeated gaps will tell you whether the first project is integration, taxonomy, staff workflow, identity matching, or outcome synchronization.