Adaptive Difficult Conversation Coach
Bridging the Gap in Managerial Communication
The Business Challenge
Unmanaged performance issues and avoided conversations are a systemic driver of team attrition. Managers account for at least 70% of the variance in team engagement, according to Gallup's foundational State of the American Manager research — and that influence hasn't diminished. Gallup's most recent global workplace data shows manager engagement itself has fallen sharply, from 31% in 2022 to 22% in 2025, the steepest decline on record. The consequences show up directly in attrition: in a 2024 Gallup study, 42% of voluntary leavers said their manager or organization could have done something to prevent their departure.
Within that pattern, this simulation targets a specific, high-risk moment: a newly promoted manager avoiding an early performance red flag until it hardens into a formal HR escalation. This simulation is a self-directed spec project built to model that failure mode — the client (Horizon Coach) and the employee scenario are illustrative.
Industry Sources:
Gallup. State of the American Manager: Analytics and Advice for Leaders. Washington, D.C.
Gallup. State of the Global Workplace: 2026 Report. Washington, D.C.
Gallup. "42% of Employee Turnover Is Preventable but Often Ignored" (Tatel & Wigert, 2024, updated 2026).
Target Audience & Needs Assessment
This simulation targets first-line managers and newly promoted team leads. Traditional passive video modules fail to build real-time communication skills for this persona. The goal is to transition these managers from passive understanding to adaptive performance—the ability to interpret ambiguity, regulate interpersonal dynamics, and co-create solutions within complex interpersonal dynamics.
Scenario Context: Symptom vs. Root Cause
Learners enter the simulation with varying levels of confidence and experiences. To prevent learners from simply guessing the “safest” answer, the simulation anchors itself in ambiguous performance evidence.
The Business Symptom: The learner receives an urgent email flagging a 28% drop in deliverable completion (metrics illustrative) over the past 30 days.
The Hidden Root Cause: By reviewing the dashboard, the learner discovers conflicting indicators: high client satisfaction, a recent department restructuring, and a massive 40% spike in logged overtime.
The Instructional Goal: A reactive manager will focus solely on the output drop and issue mandates. A diagnostic manager will identify the overtime spike as a capacity bottleneck and use empathetic diagnosis before action.
The Solution: Adaptive Conversation Simulator
Traditional branching scenarios often rely on forced, linear scripts that break the learner's sense of agency. To drive genuine behavioral transfer, I built an Adaptive Conversation Simulator. Instead of restricting learners to predetermined slide paths, the module dynamically shifts the employee’s emotional state—Guarded, Neutral, Open, Aligned, or Collaborative—based on real-time choices. This state-driven architecture allows the dialogue to adapt instantly to the manager's approach across four sequential phases: Safety, Understand, Solve, and Commitment—preceded by a strategic planning screen.
Pre-Meeting (Profile Strategy): Before the dialogue begins, the learner reviews the profile and sets their conversational intent. This choice triggers the underlying variables before a single word is spoken.
Pre-Meeting (Profile Strategy)
Phase 1 (Safety) - The Adaptive Result: The exact same opening slide branches dramatically based on the pre-meeting choice.
The Directive Path (Top): Choosing a strictly directive opener locks the Trust Index at a low 35/100, and the employee begins the conversation visibly Guarded.
The Empathetic Path (Bottom): Choosing to probe root causes elevates the Trust Index, opening a completely different conversational pathway.
The Directive Path ⬆
The Empathetic Path ⬆
From here, the learner must navigate resistance to uncover the true workload bottleneck, evaluate the employee's proposed solutions to offload administrative tasks, and finally co-create a mutual action plan.
Technical Variable Architecture (Storyline 360)
A branching scenario like this one can easily spiral into an unmanageable tangle of one-off variables, each hardcoded to a single choice, with no clean way to reuse logic across slides.
To avoid that, this simulation runs on just three core variables that calculate every outcome dynamically, keeping the system lean while still supporting open-ended branching.
Alex_State (Text): Carries the employee's emotional state across slides to trigger the correct branching dialogue. To separate front-end visuals from back-end logic, the system writes to parallel objects per choice: updating the visible character state while simultaneously setting the underlying Alex_State variable.
Relationship_Score (0–100): Modified dynamically by choice selection. It drives the live "Trust & Safety Index" progress bar, providing immediate front-end feedback.
Error_Count (Numeric): Tracks suboptimal choices behind the scenes. This trigger is strictly conditional—incrementing only if a specific suboptimal response is selected before continuing—driving the final Leadership Reflection Dashboard and the pass/retry outcome.
Technical variable Architecture
Universal Design Strategy (WCAG 2.1 Level AA)
To ensure total control over the visual hierarchy and accessibility standards, I custom-designed a clean UI component system in Figma prior to Storyline development:
Visual & State Indicators: Character state changes pair facial expression shifts with explicit text indicators (e.g., "Emotional State: GUARDED") to ensure state transitions are transparent to screen reader users.
Color Contrast: All HUD indicators, dialogue text, and body copy meet or exceed the WCAG AA 4.5:1 contrast ratio requirement against their respective backgrounds.
Keyboard Focus: Tab order strictly follows speaker context: Speaker Header → Speech Bubble → Choice 1 → Choice 2 → Choice 3.
Measurement & Evaluation Strategy
To evaluate operational ROI, a multi-tiered Kirkpatrick evaluation framework was established for the simulated deployment.
Level 2 (Competency & Recovery): Real-time evaluation culminates in a custom Leadership Reflection Dashboard. A minimum benchmark of 80/100 is required to pass—intentionally set below a strict 100% mastery model to allow a realistic misstep and course-correction rather than punitive retries.
Level 3 (Behavioural Transfer): Tracked 60–90 days post-training via automated HR conflict escalation data and voluntary team attrition rates among high-potential team members.