Adoption Friction Detector

Overview

Identify barriers to adoption before they derail rollout. The goal is to surface friction points while there's still time to address them - not after users have already resisted.

Core principle: Technical success (accuracy, pilot) doesn't guarantee adoption. Organizational friction kills more AI projects than technical failure.

Friction Categories

Every rollout faces friction across these categories:

Category	Description	Example
Trust	Belief the AI works correctly	"It will make mistakes we get blamed for"
Incentive	Misaligned rewards	Volume metrics conflict with review time
Cultural	Organizational resistance	Influential skeptics, "we tried this before"
Workflow	Process disruption	New queue types, different daily rhythm
Skill	Capability gaps	Training-doing gap, edge case uncertainty
Technical	System usability	Poor UX, slow response, unreliable

Output Format

yaml

adoption_friction_analysis:
  capability: "[Name]"
  rollout_stage: "[Pre-production | Early rollout | Stalled]"
  overall_adoption_risk: "[LOW | MEDIUM | MEDIUM-HIGH | HIGH]"
  adoption_projection:
    critical_mass_threshold: "[X%]"
    projected_immediate_adopters: "[Y%]"
    gap_to_threshold: "[Z%]"
    projection_confidence: "[HIGH | MEDIUM | LOW]"

friction_inventory:
  friction_id:
    category: "[Trust | Incentive | Cultural | Workflow | Skill | Technical]"
    description: "[What the friction is]"
    root_cause: "[Underlying issue]"
    affected_roles: ["[Role 1]", "[Role 2]"]
    affected_count: "[Number of people]"
    severity: "[HIGH | MEDIUM | LOW]"
    influence_on_adoption: "[HIGH | MEDIUM | LOW]"
    priority_score: "[Severity × Influence calculation]"

friction_cascade:
  primary_frictions: ["[Root causes]"]
  cascade_effects:
    - trigger: "[Friction A]"
      amplifies: "[Friction B]"
      mechanism: "[How A makes B worse]"
  cascade_risk: "[If X not addressed, Y and Z will worsen]"

detection_signals:
  metrics:
    - metric: "[What to measure]"
      baseline: "[Expected value]"
      warning_threshold: "[When to act]"
      indicates: "[What it means]"
  behaviors:
    - "[Observable behavior signaling friction]"
  feedback_channels:
    - channel: "[How to solicit]"
      frequency: "[How often]"

intervention_priority:
  must_do_before_rollout:
    - intervention: "[Specific action]"
      addresses: "[Friction ID]"
      owner: "[Who]"
      deadline: "[When]"
  do_during_rollout:
    - intervention: "[Specific action]"
      trigger: "[When to execute]"
  contingency:
    - if: "[Signal observed]"
      then: "[Intervention]"

stakeholder_map:
  high_influence_skeptics:
    - role: "[Who]"
      concern: "[Their specific issue]"
      approach: "[How to engage]"
  champions:
    - role: "[Who]"
      leverage: "[How to use them]"
  silent_majority:
    estimated_size: "[X%]"
    likely_behavior: "[Follow leaders | Wait and see | Passive resist]"

adoption_trajectory:
  week_1: "[Expected % adoption]"
  week_2: "[Expected % adoption]"
  week_4: "[Expected % adoption]"
  warning_milestones:
    - week: "[N]"
      minimum_adoption: "[X%]"
      if_below: "[Action]"

rollback_criteria:
  adoption_floor: "[Below X%, abort]"
  quality_floor: "[Below Y accuracy, abort]"
  decision_authority: "[Who decides]"
  rollback_trigger: "[Specific condition]"

post_rollout_monitoring:
  ongoing_metrics:
    - metric: "[What to track]"
      frequency: "[How often]"
      owner: "[Who monitors]"
  friction_resolution_criteria:
    - friction: "[Friction ID]"
      resolved_when: "[Condition]"
  long_term_health_check:
    frequency: "[Monthly | Quarterly]"
    includes: "[What to assess]"

Friction Cascade Analysis

Frictions don't exist in isolation. Map how they amplify each other:

yaml

friction_cascade:
  primary_frictions:
    - "incentive_misalignment"  # Root cause
    - "previous_failure_shadow"  # Root cause

  cascade_effects:
    - trigger: "incentive_misalignment"
      amplifies: "trust_deficit"
      mechanism: |
        Staff rushing to meet volume metrics →
        More errors → "See, AI doesn't work" →
        Lower trust in system

    - trigger: "previous_failure_shadow"
      amplifies: "skeptic_influence"
      mechanism: |
        "We tried this before" narrative →
        Validates skeptic concerns →
        Skeptics gain credibility

    - trigger: "layoff_anxiety"
      amplifies: "passive_resistance"
      mechanism: |
        "If we make this work, we lose jobs" →
        Subtle non-adoption →
        Adoption metrics lag

  cascade_risk: |
    If incentive misalignment not fixed before rollout,
    trust deficit will compound with skeptic influence,
    making post-rollout recovery much harder.

  intervention_sequence:
    1: "Fix incentives first (root cause)"
    2: "Address skeptics (high influence)"
    3: "Layoff commitment (anxiety reduction)"
    4: "Previous failure acknowledgment (trust building)"

Intervention Priority Matrix

Calculate priority from severity and influence:

yaml

intervention_priority:
  scoring:
    severity:
      HIGH: 3
      MEDIUM: 2
      LOW: 1
    influence_on_adoption:
      HIGH: 3
      MEDIUM: 2
      LOW: 1
    priority_score: "severity × influence"

  prioritized_list:
    - friction: "incentive_misalignment"
      severity: HIGH (3)
      influence: HIGH (3)
      priority: 9
      intervention: "Revise metrics before rollout"

    - friction: "vocal_skeptics"
      severity: HIGH (3)
      influence: HIGH (3)
      priority: 9
      intervention: "One-on-one engagement, accountability model"

    - friction: "layoff_anxiety"
      severity: MEDIUM (2)
      influence: MEDIUM (2)
      priority: 4
      intervention: "Written no-layoff commitment"

  resource_allocation: |
    Priority 9: Must address before rollout
    Priority 6-8: Address in Week 1
    Priority 4-5: Monitor and address if signals appear
    Priority <4: Standard rollout process

Early Adopter Strategy

Leverage champions systematically:

yaml

early_adopter_strategy:
  identification_criteria:
    - "Participated in pilot with positive experience"
    - "Expressed interest or asked questions in training"
    - "Generally open to new tools (history)"
    - "Respected by peers (influence)"

  leverage_plan:
    visibility:
      - "Include in rollout communications"
      - "Ask for testimonials after Week 1"
      - "Recognize publicly for early success"

    peer_influence:
      - "Pair with hesitant adopters"
      - "Assign as shift/team resource"
      - "Invite to debrief sessions to share experience"

    feedback_loop:
      - "Weekly check-in for issues"
      - "Channel for suggestions"
      - "Recognition for catches/improvements"

  target: "20% enthusiastic adopters → influence middle 60%"

Rollback Criteria

Define when to abort:

yaml

rollback_criteria:
  adoption_floor:
    threshold: "50% active usage by Week 4"
    if_below: "Pause rollout, intensive intervention"

  quality_floor:
    threshold: "Accuracy drops below 94%"
    if_below: "Rollback to previous process"

  friction_escalation:
    threshold: "Union grievance filed or escalation to HR"
    if_triggered: "Pause rollout, formal review"

  decision_authority:
    pause: "Operations Manager"
    rollback: "Operations VP + Risk approval"

  rollback_process:
    1: "Announce pause/rollback clearly"
    2: "Revert to previous process"
    3: "Post-mortem on friction causes"
    4: "Plan for re-attempt with fixes"

Common Friction Patterns

Financial Services Specific

Pattern	Why It Happens	Mitigation
Blame anxiety	Audit trails make errors visible	Explicit accountability model
Compliance caution	Fear of regulatory issues	Compliance sign-off before rollout
Override reluctance	"What if I'm wrong?"	Clear override criteria, no blame for good-faith overrides
Volume pressure	Settlement deadlines conflict with review time	SLA adjustment for review queue

Union Considerations

yaml

union_friction:
  concerns:
    - "AI will reduce headcount"
    - "Workload changes without negotiation"
    - "New performance metrics"

  mitigation:
    before_rollout:
      - "Formal union briefing with documentation"
      - "Written no-layoff commitment (time-bound)"
      - "Workload study results shared"

    during_rollout:
      - "Regular union liaison check-ins"
      - "Immediate escalation path for concerns"

    documentation:
      - "Meeting minutes with union"
      - "Written commitments on file"
      - "Grievance prevention log"

Common Mistakes

Mistake	Why It's Wrong	Do This Instead
One friction analysis	Frictions compound	Map cascades explicitly
Equal treatment	Not all frictions matter equally	Priority matrix (severity × influence)
Rollout-focused only	Adoption is ongoing	Post-rollout monitoring plan
Generic mitigations	"Communicate more" doesn't work	Specific, actionable interventions
Ignore skeptics	Skeptics influence silent majority	Direct engagement, accountability model
Volume metrics unchanged	Correct behavior punished	Fix incentives before rollout

Red Flags in Your Output

If your analysis has these, it's not ready:

•No severity ratings (all frictions treated equally)
•No cascade analysis (frictions listed independently)
•No intervention priority (no resource allocation guidance)
•No rollback criteria (no fail-safe)
•Generic mitigations ("increase communication")
•Missing early adopter strategy
•No post-rollout monitoring

Financial Services Context

Financial services adoption requires:

Compliance Integration

•Regulatory training requirements may affect rollout timing
•Audit trail concerns are legitimate (address accountability)
•Override documentation is regulatory requirement, not bureaucracy

Incentive Structures

•Often volume-based, conflicts with quality-focused AI work
•Must address before rollout, not after

Union Relationships

•Formal engagement required for AI deployments
•Written commitments on headcount
•Escalation paths for concerns

Friction Analysis Checklist

Before finalizing: