tracking-gaps

Skill: Tracking Gap Identification

Purpose

Assess whether the data needed for an analysis actually exists, identify what's missing, and produce prioritized instrumentation requests for engineering when gaps are found.

When to Use

Apply this skill after the Data Explorer agent inventories available data, when an analysis requires data that might not exist, or when initial query results suggest incomplete tracking. Run before committing to an analysis approach.

Instructions

Gap Detection Process

Define requirements (Step 1) before inventorying what exists; starting from "what we have" is how gaps get missed.

Step 1: Define Data Requirements

Before checking what data exists, list every data point the analysis NEEDS. This ensures you don't miss gaps.

For each analytical question, create a structured requirements table:

| Requirement | Needed For | Granularity | Time Range |
|-------------|-----------|-------------|------------|
| [event/field] | [which analysis step] | [per user/session/event] | [last 30d, 90d, etc.] |

Step 2: Inventory Available Data

Map each requirement to what actually exists:

| Requirement | Status | Source | Notes |
|-------------|--------|--------|-------|
| [event/field] | AVAILABLE / PARTIAL / MISSING / DERIVABLE | [table.column] | [caveats] |

Status definitions:

• AVAILABLE: Data exists, is clean, and covers the needed time range
• PARTIAL: Data exists but has gaps — missing time ranges, incomplete segments, or quality issues
• MISSING: Data is not tracked at all — requires new instrumentation
• DERIVABLE: Data doesn't exist directly but can be approximated from other available data

Step 3: Design Workarounds for Gaps

For each PARTIAL or MISSING item, evaluate workarounds:

### Gap: [what's missing]
**Impact on analysis:** [how this gap affects what we can conclude]
**Workaround:** [how to approximate using available data]
**Confidence with workaround:** [High/Medium/Low]
**Workaround limitations:** [what the approximation gets wrong]