← Index

ROI Measurement Framework

apc/gwyn // 3 foundational metrics → portfolio dashboard → AFA pricing → client reporting

The JD asks Charlie to "develop frameworks to track AI solution performance" and "maintain a portfolio dashboard highlighting adoption, client outcomes, and return on innovation." Most candidates can talk about ROI. Charlie built the measurement system. Three foundational metrics, tracked from day one, feeding decisions that partners actually care about.

The AI ROI blind spot in legal:

85%
of firms don't track AI ROI
18%
of legal professionals say their org measures returns
68%
of in-house counsel don't know if outside firms use AI

APC is trying to be in the 18%. The K&I Program Manager builds the capability that gets them there.

      graph LR
        subgraph METRICS["Three Foundational Metrics"]
          M1["1. Adoption
% team logging in
% active matters
using AI"]:::m1
          M2["2. Time Saved
hours per task
before vs. after AI
per matter, per client"]:::m2
          M3["3. Profitability
total cost vs. revenue
write-downs tracked
margin per matter"]:::m3
        end

        subgraph DASH["Portfolio Dashboard"]
          PD["Adoption · Outcomes
Cost · ROI per agent
Firm-wide view"]:::dashboard
        end

        subgraph DECISIONS["Decisions the Data Enables"]
          AFA["AFA Pricing
flat fees priced
with margin confidence
not guesswork"]:::afa
          CR["Client Reporting
quarterly impact:
time saved, accuracy,
scope of AI use"]:::client
          EVAL["Evaluation Engine
hard veto on safety/bias
continuous drift detection
audit trail proves claims"]:::eval
        end

        M1 --> PD
        M2 --> PD
        M3 --> PD
        PD --> AFA
        PD --> CR
        PD --> EVAL

        classDef m1 fill:#241a35,stroke:#a78bfa,color:#e8e0f0,stroke-width:1.5px
        classDef m2 fill:#1a2e3d,stroke:#22d3ee,color:#cffafe,stroke-width:1.5px
        classDef m3 fill:#1a2a1a,stroke:#4ade80,color:#dcfce7,stroke-width:1.5px
        classDef dashboard fill:#1a1830,stroke:#a78bfa,color:#ede0ff,stroke-width:2px
        classDef afa fill:#2a2618,stroke:#fbbf24,color:#fef3c7,stroke-width:1.5px
        classDef client fill:#1a2a1a,stroke:#4ade80,color:#dcfce7,stroke-width:1.5px
        classDef eval fill:#2a1a18,stroke:#fb923c,color:#fde8d0,stroke-width:1.5px

        classDef subg fill:transparent,stroke:#3a3050,stroke-dasharray:6 4,color:#9b8fb8
        class METRICS,DASH,DECISIONS subg
    

Three Foundational Metrics

01

Adoption / Utilization

You can't evaluate value without understanding use. Measured from day one, per agent, per practice group.

  • % of team logging in daily
  • % of active matters using the tool
  • Repeat usage rate (single-use vs. sustained)
  • Shadow IT displacement tracking
02

Time Saved

Quantifies efficiency. Supports AFA pricing. Satisfies ABA 512's reasonable-fees requirement with data, not narrative.

  • Hours per task before vs. after AI
  • Per-matter, per-client granularity
  • Segmented by task type (review, drafting, research)
  • Feeds directly into AFA pricing models
03

Profitability per Matter

Determines whether efficiency equals margin or just speed. Speaks the language partners actually care about.

  • Total cost vs. revenue per matter
  • Write-downs tracked and attributed
  • Realization rate with vs. without AI
  • PEP and RPL impact projections

What the Data Enables

Portfolio Dashboard

Every agent: cost tracking, operational status, adoption metrics, outcome metrics, evaluation scores. Leadership sees what's deployed, who's using it, and what it's saving. The adoption roadmap synthesizes into a prioritized sequence.

AFA Pricing Confidence

If a fixed-fee due diligence engagement takes 2 AI-assisted hours instead of 15, the firm can price that flat fee with margin. Without the data, it's guessing. Most firms are guessing. 85% of them.

Client Reporting

68% of in-house counsel don't know if outside firms use AI. 85% say firms should disclose AI use. The firm that communicates clearly — and proves value with data — wins the client conversation.

Evaluation Engine

Every agent scored on safety, bias, accuracy, functionality — before deployment and continuously after. Hard veto rule. The audit trail that proves to a client the AI is working as claimed. Not governance theater. System output.

The Billable Hour Tension — and the Resolution

Partners profit from the spread on associate hours. AI compresses associate hours. Every efficiency gain is simultaneously a win for the client and a revenue question for the firm. Pretending this tension doesn't exist is how AI programs lose partner support.

The resolution: use AI-driven efficiency to make AFAs profitable. ~90% of legal spend still flows through hourly arrangements. AFA revenue is flat at ~23.5%. The shift is year 3 of a 10-15 year transition. The firms that track AI ROI now are the ones positioned to price AFAs with confidence when the market tips. The firms that don't track — they'll be guessing when the client demands a fixed fee.

The competence pipeline dimension: tracking time saved isn't just about pricing. It's about proving that junior lawyers are still developing judgment even when AI handles the grunt work. If first-years never do first-pass review, they never develop the pattern recognition to spot AI errors. The metrics framework has to measure learning, not just efficiency — because the profession is profiting today by mortgaging the expertise of tomorrow's partners.

The K&I Program Manager's ROI mandate: the JD asks Charlie to "develop frameworks to track AI solution performance." That's not a reporting function. It's an operational capability — instrumented before launch, feeding decisions from day one, proving value with data partners trust and clients demand. Track the three foundational metrics. Feed the portfolio dashboard. Enable AFA pricing confidence. Deliver client reports that prove the value. 85% of firms can't do this. APC can.