AI Skill Report Card

Planning Product Strategy

A91·Sep 27, 2026·Source: Web
Markdown
--- name: planning-product-strategy description: Develops product vision, roadmaps, feature prioritization matrices, and execution plans from user needs and business goals. Use when defining product direction, prioritizing a feature backlog, planning a roadmap, or translating user needs into an execution plan. --- # Product Strategy Planning
15 / 15

Given a product context (target users, business goals, known pain points, competitive landscape), produce these five artifacts in order:

  1. Vision — one paragraph, outcome-focused
  2. Roadmap — time-horizon buckets (Now / Next / Later)
  3. Feature Matrix — features scored against value & effort
  4. Priority Score — ranked backlog using a weighted formula
  5. Execution Plan — owners, milestones, dependencies for top priorities

Do not skip straight to features. Vision and user needs must anchor every downstream artifact — if a feature can't be traced back to the vision or a validated need, flag it instead of silently including it.

Recommendation▾
Add a third example showing a poor/anti-pattern output (e.g., feature-first list without vision) to reinforce contrast between good and bad outcomes
14 / 15

Progress:

  • Step 1: Clarify user needs and business goals (ask only if truly missing; otherwise infer from context and state assumptions)
  • Step 2: Draft Vision statement
  • Step 3: Build Roadmap (Now/Next/Later, not fixed dates unless given)
  • Step 4: Build Feature Matrix (Value vs Effort vs Confidence)
  • Step 5: Compute Priority Score per feature
  • Step 6: Write Execution Plan for top 3–5 priorities

Step 1: Clarify inputs

Identify: target user segment, core problem, business objective (growth/retention/revenue/efficiency), constraints (timeline, team size, tech debt). If unspecified, state explicit assumptions rather than pausing the whole output.

Step 2: Vision

One paragraph. Format: "For [user], who [problem], [product] is a [category] that [key benefit]. Unlike [alternative], we [differentiator]."

Step 3: Roadmap

Three horizons, not calendar dates:

  • Now (in progress / next sprint-quarter): committed, scoped
  • Next (1–2 quarters out): validated direction, not yet scoped
  • Later (exploratory): bets and hypotheses

Each item: one-line outcome (not a feature name). E.g., "Reduce onboarding drop-off by 20%" not "Build new onboarding flow."

Step 4: Feature Matrix

Table with columns: Feature | User Value (1-5) | Business Value (1-5) | Effort (1-5, 5=hardest) | Confidence (Low/Med/High) | Notes

Step 5: Priority Score

Default formula (RICE-inspired, adapt if inputs given):

Priority Score = (User Value + Business Value) × Confidence Multiplier / Effort

Confidence Multiplier: High=1.0, Med=0.7, Low=0.4

Rank descending. Show the math, not just the final number.

Step 6: Execution Plan

For each top-ranked feature: Owner (role, not name unless given) | Milestone breakdown (2-4 steps) | Dependencies/risks | Success metric (leading indicator tied back to Vision)

Recommendation▾
Include guidance on handling conflicting stakeholder priorities or multiple business goals competing for the same roadmap slot
18 / 20

Example 1: Input: "MusGo is a music collaboration app for indie artists. Users struggle to find collaborators and manage version conflicts on shared tracks. Goal: increase weekly active collaboration sessions."

Output:

Vision: For indie musicians who struggle to find and coordinate with collaborators,
MusGo is a collaboration platform that makes co-creating music as frictionless as
messaging. Unlike generic file-sharing tools, we handle version control and
discovery natively for music production.

Roadmap:
- Now: Ship version-conflict resolution for shared tracks
- Next: Launch collaborator discovery/matching
- Later: Real-time co-editing sessions

Feature Matrix:
| Feature                    | UserVal | BizVal | Effort | Confidence | Notes |
|---------------------------|---------|--------|--------|------------|-------|
| Version conflict resolver | 5       | 4      | 3      | High       | Core pain point |
| Collaborator matching     | 4       | 5      | 4      | Med        | Needs matching algo |
| Real-time co-edit         | 3       | 3      | 5      | Low        | High tech risk |

Priority Score:
- Version conflict resolver: (5+4)×1.0/3 = 3.0 → #1
- Collaborator matching: (4+5)×0.7/4 = 1.58 → #2
- Real-time co-edit: (3+3)×0.4/5 = 0.48 → #3

Execution Plan (Version conflict resolver):
- Owner: Eng lead (backend) + PM
- Milestones: (1) design conflict-detection logic (2) build merge UI
  (3) beta with 20 power users (4) full rollout
- Dependencies: requires track storage schema update
- Success metric: % of shared tracks with unresolved conflicts, target <5%

Example 2: Input: "prioritize: dark mode, export to Spotify, offline mode" (no context given)

Output: State assumption ("Assuming general productivity/consumer app, no stated business goal — scoring on user value and typical effort only"), then produce Feature Matrix + Priority Score only, and note that Vision/Roadmap/Execution Plan need more context to be meaningful — offer a best-effort version anyway rather than blocking.

Recommendation▾
Consider adding a brief note on how to handle re-prioritization when new information arrives mid-roadmap
  • Always tie features back to the Vision; if you can't, say so explicitly instead of forcing a justification.
  • Prefer outcome language ("reduce churn by X") over output language ("build feature Y") in Vision and Roadmap.
  • Keep Feature Matrix scores relative and consistent — recalibrate if scores cluster all at 4-5.
  • Show scoring math for Priority Score; unexplained rankings erode trust.
  • Keep Execution Plans to top 3-5 items — a plan for everything is a plan for nothing.
  • State assumptions explicitly when inputs are incomplete; don't block on missing info.
  • Don't jump straight to a feature list without a Vision — it produces a backlog, not a strategy.
  • Don't use fixed calendar dates in early-stage roadmaps unless the user provided them; horizons (Now/Next/Later) age better.
  • Don't let Effort scores be guessed by feel alone — flag when Effort is uncertain rather than presenting false precision.
  • Don't average scores when Confidence is Low — surface the uncertainty instead of hiding it in the final number.
  • Don't include every requested feature in the Execution Plan; that's what the backlog/matrix is for.
0
Grade AAI Skill Framework
Scorecard
Criteria Breakdown
Quick Start
15/15
Workflow
14/15
Examples
18/20
Completeness
15/20
Format
15/15
Conciseness
14/15