Back to Case Studies

Designing a Fair, Transparent Fact-Checking Assignment System

Policy Research Organization Workload Governance

Every publication goes through fact-checking, and every assignment matters. The result was a system that makes fair, defensible reviewer assignments automatically, in seconds.

Results at a Glance

< 5 sec

Assignment Time

98%

Faster

120+ hrs

Saved / Year

The Challenge

Before every assignment, the editor had to pause and weigh: Who reviewed most recently? Who has carried more load? Who declined last week? Is there a conflict? Fairness lived entirely in one person's head.

  • Fairness depended on one person's memory, not a system
  • A cognitive bandwidth problem, not a reviewer problem
  • A high-stakes environment that required defensible, auditable decisions

The Approach

I designed a fairness model built on two rules: eligibility (recency) and priority (workload).

  • Dynamic median logic recalculates eligibility daily, rather than a fixed "30 days"
  • Automatic conflict filtering for authors, collaborators, and project managers
  • Cooldown logic for recent declines and protected windows

Salesforce Objects & Tools

Eligibility & priority rules Dynamic median calculation Conflict-of-interest filtering Cooldown & window enforcement

Results

  • Reduced assignment deliberation from 3–5 minutes per publication to under 5 seconds (98% faster), across 200+ fact-checking cycles per year
  • Achieved 100% defensible, auditable decisions with zero bias claims or reviewer complaints since go-live
  • Freed up 120+ hours per year of senior leadership time previously spent mentally simulating fairness and workload equity

FAQ

Have a Salesforce system that needs this kind of work?

Let's talk about how I can bring the same clarity, stability, and structured thinking to your organization's Salesforce.