General analytics
AuraScore 81/100

Multimodal Prompt Refinement and Canvas Engagement Analytics Brief

Analyze multimodal prompt iteration loops, inpainting churn, and canvas tool utilization metrics across user tiers.

Use this template when investigating product engagement, drop-off during prompt re-rolls, and interaction depth inside generative image canvas interfaces. It outputs a comprehensive product analytics brief focused on iteration velocity and feature adoption.

Template

Role: Staff Product Analytics Director for Generative Platforms with expertise in behavioral telemetry and creative workflow instrumentation.

Context

  • User Tier Cohort: {{platform_tier}}
  • Target User Archetype: {{user_cohort_definition}}
  • Canvas Editing Capabilities: {{canvas_toolset_features}}
  • Longitudinal Study Period: {{target_retention_window}}
  • Primary Friction Metric: {{regeneration_frequency_metric}}
  • Success Conversion Event: {{export_conversion_threshold}}

Task

Synthesize multimodal canvas interaction logs into a high-level product analytics brief that diagnoses prompt refinement fatigue, evaluates mask-and-inpaint feature efficiency, and models user retention drivers.

Method

  1. Map the end-to-end user journey from initial text-to-image prompt submission to final asset export.
  2. Quantify prompt modification patterns (re-weighting, keyword stripping, style modifiers) across {{platform_tier}}.
  3. Measure session abandonment rates correlated with high values of {{regeneration_frequency_metric}}.
  4. Analyze adoption curves and task completion times for specific tools within {{canvas_toolset_features}}.
  5. Segment time-to-value milestones for {{user_cohort_definition}} comparing linear prompting against canvas inpainting.
  6. Determine the statistical inflection point where prompt iterations transition from creative exploration into churn.
  7. Model downstream retention over {{target_retention_window}} as a function of reaching {{export_conversion_threshold}}.

Constraints

  • MUST calculate correlation coefficients between regeneration fatigue and subsequent churn.
  • MUST evaluate feature engagement specifically across the tools listed in {{canvas_toolset_features}}.
  • MUST NOT suggest user interface redesigns without direct telemetry-backed behavioral evidence.
  • All conversion funnel steps MUST display absolute user volume, step-over-step drop-off, and cohort relative rates.

Output format

An analytics brief comprising:

  • Funnel Health Overview (Text Prompt -> Inpaint -> Canvas Upscale -> {{export_conversion_threshold}})
  • Prompt Churn & Fatigue Analysis (Statistical breakdown of {{regeneration_frequency_metric}})
  • Toolset Engagement Matrix (Utilization, Average Dwell Time, Success Rate for {{canvas_toolset_features}})
  • Cohort Retention Modeling (Impact across {{target_retention_window}})
  • High-Leverage Product Recommendations (4 prioritized product interventions with expected impact)

Self-review

  • Confirm all funnel stages directly terminate at {{export_conversion_threshold}}.
  • Verify that interaction metrics are segmented by the designated {{platform_tier}}.
  • Check that recommendations directly target reducing the friction measured in {{regeneration_frequency_metric}}.
AuraScore breakdown
81/100Provisional
Instruction clarity15/15 · Strong

Explicit role, a named task, and discrete steps the model can follow.

Context architecture12/12 · Strong

Background, inputs and variables the model needs before it starts.

Constraint engineering12/12 · Strong

Hard boundaries — what the model must and must not do.

Output specification6/14 · Thin

A named, field-level shape for the response.

Reasoning structure10/10 · Strong

Ordered work items that force analysis before an answer.

Model compatibility10/10 · Strong

Length and structure that travel across frontier models.

Token efficiency5/10 · Thin

Signal density — instruction weight without padding.

Reusability7/7 · Strong

Documented variables so the scaffold adapts to new inputs.

Robustness3/5 · Adequate

Quality bar, assumptions and behaviour when inputs are thin.

Observed performance1/5 · Thin

How much real usage the template has behind it.

data-analytics
data-general
image-multimodal-prompting
product-analytics
multimodal-canvas
prompt-telemetry