Owner Olivia – Discovery Sheet
Related Jobs-To-Be-Done
JTBD‑2 — Detect and Diagnose Issues Quickly
When my service begins to degrade or fail
I want to understand what’s happening and why using correlated signals
so I can restore service health quickly and reduce customer impact.
1. Hypothesis Statement
We believe that Owner Olivia needs a single, correlated view of logs, metrics, and traces during incidents, so that she can diagnose issues quickly without jumping between tools or guessing root cause. We’ll know this is true when she consistently uses the correlated incident view during triage and resolves issues faster than with her current workflow.
2. Top 3 Riskiest Assumptions
| Assumption | Type | Risk Reason |
|---|---|---|
| Olivia prefers correlated, incident‑focused insights over traditional dashboards | Value | If not true, the feature will not solve her biggest pain |
| We can reliably stitch logs, metrics, and traces together in real time | Feasibility | Requires strong data quality and OTel consistency |
| Incident signals can be distilled into a clear, low‑noise experience | Usability | Too much data or noise will overwhelm users |
3. Lean Experiment (Single Test)
Experiment Type: High‑touch usability test using a manually prepared “correlated incident snapshot” during active or simulated incidents.
Method:
- Select 2–3 real incidents from the past 30 days.
- Manually compile a single view showing: • timeline of events • trace exemplar • key log lines • error-rate and latency graphs
- During an incident simulation, ask Olivia to troubleshoot using only the snapshot.
- Measure time to understanding and perceived clarity.
What It Tests:
- Whether correlated context shortens debugging time (Value)
- Whether the format is intuitive and trusted (Usability)
- Whether data can be extracted and stitched manually today (Feasibility)
Success Signal:
- Olivia identifies root cause significantly faster
- She expresses high confidence in the unified view
- She stops switching tools during the exercise
Failure Signal:
- She still needs dashboards or raw logs
- Snapshot is perceived as unclear or incomplete
- She distrusts the correlations or signals
4. Evidence Summary
Observed Feedback (simulated expected): “This is exactly what I need during an incident. It saves me from clicking through five tools just to figure out what happened.”
Interpretation: Olivia values clarity, correlation, and a single place to start debugging more than detailed dashboards.
Learning: The real need is a Correlated Incident View—a guided, consolidated snapshot of everything that matters in the first 5 minutes of an incident.
Reframed JTBD: “Help me understand what’s happening during an incident without searching through multiple tools.”
5. Decision
Outcome: Hypothesis validated enough to proceed.**
Decision: Move forward with designing & building a Correlated Incident View (MVP).
Reasoning:
- Strong qualitative signal that this reduces cognitive load
- Matches Olivia’s top frustrations (tool‑sprawl, slow diagnosis, missing context)
- Data availability seems sufficient for an MVP
- High value and high adoption probability
Next Step: Create Feature Issue: “Correlated Incident View (MVP)”