| name | influential-factors-detective |
| description | Identify and analyze qualitative factors that influence KPIs but can't be measured as numbers. Use when users need to find non-quantifiable drivers affecting metrics (like copy quality, design aesthetics, timing), plan A/B tests for subjective elements, bridge the gap between data and design/content decisions, understand why quantitative analysis is incomplete, optimize messaging or UX elements, or identify the actual levers to pull for improving KPIs at the leaf nodes of a KPI tree. |
Influential Factors Detective
Systematically identify and test qualitative factors that drive KPI performance but can't be directly measured.
Core Concept
Influential Factors = Elements you can change that affect KPIs, but whose "quality" can't be directly measured—only their impact on downstream metrics.
Location in KPI Tree:
North Star Metric
├─ Quantifiable KPI
│ └─ Quantifiable KPI
│ └─ 🔍 Influential Factor (qualitative leaf node)
Key insight: Quantitative analysis tells you WHERE the problem is. Influential factors tell you WHAT to change.
When to Use This Skill
Trigger this skill when:
- User asks to identify influential factors for a KPI
- KPI changed but quantitative drivers don't explain why
- User needs to optimize subjective elements (copy, design, UX)
- User wants to plan A/B tests
- User asks "what can we change?" after reaching KPI tree leaf nodes
- User needs to bridge analytics and creative decisions
Six Categories of Influential Factors
1. Copy & Messaging
Word choice, tone, length, clarity, emotional appeal, value prop framing, CTA phrasing, urgency language
Example: Email subject line—can't measure "quality" but can test which one drives higher open rates
2. Visual Design
Layout, colors, typography, visual hierarchy, images, icons, button design, white space
Example: CTA button color—can't measure "effectiveness" but can test which color drives more clicks
3. User Experience
Navigation, form fields, progress indicators, error messages, loading states, micro-interactions
Example: Form length—can't measure "optimal length" but can test 5 vs 10 vs 15 fields
4. Timing & Frequency
Time of day, day of week, contact frequency, delay periods, pacing, response expectations
Example: Email send time—can't measure "best time" universally but can test 8am vs 12pm vs 6pm
5. Content & Media
Format (image/video/text), storytelling approach, depth, technical level, personalization
Example: Product demo format—can't measure "best format" but can test video vs screenshots vs interactive
6. Positioning & Placement
Above/below fold, sidebar vs main content, modal vs banner, navigation placement, feature prominence
Example: CTA placement—can't measure "ideal position" but can test top vs bottom vs sticky
Four Methods to Identify Influential Factors
Method 1: KPI Tree Leaf Analysis
When: Working with an existing KPI tree
Process:
- Navigate to bottom-level quantifiable KPIs
- Ask: "What elements can I change that affect this metric?"
- List all changeable but non-measurable elements
- Document as 🔍 influential factors
Template:
For KPI: [Name]
Influential Factors:
1. [Factor] - [How it affects KPI] - [Current state] - [Alternatives]
2. [Factor] - [How it affects KPI] - [Current state] - [Alternatives]
Method 2: Comparison Analysis
When: Two similar things perform differently
Process:
- Identify comparable elements (Feature A vs B, Page X vs Y)
- List quantitative differences (traffic, engagement)
- List qualitative differences (copy, design, placement)
- Hypothesize which qualitative differences drive performance gap
Example:
Landing Page A: 15% conversion | Landing Page B: 8% conversion
Qualitative differences:
- Headline (benefit vs feature)
- Hero image (people vs product)
- CTA color, Form length, Social proof placement
→ Test these factors systematically
Method 3: User Research Insights
Sources: Interviews, usability tests, session recordings, support tickets, surveys
Process:
- Collect qualitative feedback ("confusing," "overwhelming," "unclear")
- Map feedback to potential factors
- Form hypotheses about improvements
Example:
Feedback: "Wasn't sure if plan included X feature"
→ Factors: Feature list completeness, description clarity, comparison design
→ Hypothesis: Adding use case examples will increase conversion
Method 4: Best Practice Benchmarking
Process:
- Research industry best practices
- Compare your implementation
- Identify gaps as testable factors
Four Testing Approaches
A/B Testing (sufficient traffic)
- Test one factor at a time
- Create control + variant(s)
- Run to statistical significance
- Document and implement winner
Sequential Testing (low traffic)
- Measure baseline (2 weeks)
- Make change
- Measure new performance (2 weeks)
- Compare accounting for seasonality
Multivariate Testing (high traffic, multiple factors)
- Test combinations of factors
- Requires 8x+ traffic vs simple A/B
- Reveals interaction effects
- Use sparingly
Qualitative Testing (early exploration)
- Usability testing (5-10 users)
- Preference testing (side-by-side)
- 5-second tests (recall measurement)
- Inform quantitative test priorities
Workflow
Initial Assessment
- Ask user about their KPI and context
- Determine which method(s) to use for identifying factors
- Clarify: leaf node analysis, comparison, user feedback, or benchmarking?
Factor Identification
- Apply chosen method(s)
- List all potential influential factors
- Categorize by type (copy, design, UX, timing, content, placement)
- For each factor, document:
- How it could affect the KPI
- Current state
- Alternative options to test
Prioritization
Use Impact/Effort matrix:
- High Impact, Low Effort → Quick wins, test first
- High Impact, High Effort → Big bets, plan carefully
- Low Impact, Low Effort → Fill-ins, test if easy
- Low Impact, High Effort → Avoid
Score = Estimated Impact (1-5) / Effort to Test (1-5)
Test Planning
For top-priority factors:
- Choose testing approach (A/B, sequential, multivariate, qualitative)
- Define success metrics and guardrails
- Design variations
- Plan implementation and documentation
Documentation
Provide factor documentation using template in references/templates.md
Output Format
# Influential Factors for [KPI Name]
## Analysis Method(s) Used
[KPI Tree Leaf Analysis / Comparison / User Research / Benchmarking]
## Identified Factors
### Category: [Copy & Messaging / Visual Design / etc.]
#### Factor 1: [Name]
- **Parent KPI:** [Which KPI this affects]
- **Description:** [What this factor is]
- **Current State:** [How it's currently implemented]
- **Hypothesis:** [How changing this could improve the KPI]
- **Test Variations:**
- Control: [Current]
- Variant A: [Option 1]
- Variant B: [Option 2]
[Repeat for each factor]
## Prioritization
| Factor | Category | Impact (1-5) | Effort (1-5) | Priority Score | Recommendation |
|--------|----------|--------------|--------------|----------------|----------------|
| [Name] | [Category] | [#] | [#] | [Score] | [Quick Win/Big Bet/etc.] |
## Recommended Testing Sequence
1. **[Factor Name]** - [Testing approach] - [Why first]
2. **[Factor Name]** - [Testing approach] - [Why second]
3. **[Factor Name]** - [Testing approach] - [Why third]
## Next Steps
1. [Specific action]
2. [Specific action]
3. [Specific action]
Common Patterns by Context
Email Marketing: Subject lines, from name, preview text, send time, content format, CTA count
Landing Pages: Headlines, hero images, social proof placement, form fields, trust signals
In-Product: Onboarding steps, empty states, feature discovery, notification frequency
Checkout Flows: Form fields, error messages, trust indicators, guest checkout, progress indication
See references/common-factors.md for comprehensive lists by context.
Key Reminders
- Influential factors are the LEAF NODES of your KPI tree
- You can't measure their "quality" directly—only their impact on metrics
- Test systematically: one factor at a time (unless multivariate with sufficient traffic)
- Document all tests: winners, losers, and learnings
- Build institutional knowledge over time
- Target 2-4 tests per month minimum
- Win rate >40% indicates good hypothesis generation
Reference Materials
references/templates.md - Documentation templates for factors and test logs
references/common-factors.md - Comprehensive lists of factors by business context
references/case-studies.md - Detailed examples with results and learnings
Integration with Other Skills
Prerequisites:
- kpi-tree-architect - Provides the KPI tree structure where factors live as leaf nodes
- Work with KPI Tree Architect to identify where in the tree to add 🔍 influential factors
Complementary:
- Use root cause analysis to identify which factors to test when KPIs change
- Use A/B testing advisors for proper test design
- Use dashboard designers to track test results
Quality Checklist
Before testing an influential factor:
Factor Definition:
Test Design:
Implementation: