Roadmap Planning
When to Use
Activate when a founder or PM has a set of product ideas, feature requests, or strategic bets and needs to organize them into a coherent roadmap. This includes situations like "help me plan our Q2 roadmap," "prioritize this feature backlog," "what should we build next," or "create a roadmap for the next 6 months." Also activate when an existing roadmap needs restructuring from output-focused (feature lists) to outcome-focused (customer and business impact).
Context Required
- From startup-context: company stage, team size and composition, current product state, business model, key metrics, strategic goals, company OKRs.
- From the user: current roadmap or list of candidate initiatives, known customer needs, resource constraints, hard deadlines, strategy documents or company objectives for alignment, and any committed work already in progress.
Workflow
- Gather the current state -- Collect the existing roadmap or feature list. If the user provides strategy documents or company objectives, review them to understand how the roadmap should align with broader goals.
- Transform outputs to outcomes -- For each initiative, ask: What customer problem are we solving? What business metric will improve? Is there a better way to achieve the same outcome? Rewrite each item as an outcome statement.
- Apply outcome statement format -- Use: "Enable [customer segment] to [desired customer outcome] so that [business impact]." Every roadmap item must pass this test.
- Score and prioritize -- Apply RICE scoring (Reach, Impact, Confidence, Effort) to rank initiatives. Adjust for strategic alignment with company objectives and OKRs.
- Identify dependencies and sequencing -- Map which initiatives depend on others. Note technical, data, design, and business dependencies.
- Sequence into time horizons -- Place initiatives into Now (0-6 weeks), Next (6-12 weeks), and Later (12+ weeks). Use flexible release windows (quarters, not specific dates) for Later items.
- Validate capacity -- Cross-check the plan against team capacity. A roadmap that requires 3x your engineering bandwidth is a wishlist, not a plan.
- Add strategic context -- Document how outcomes align with company strategy, key assumptions about customer needs, and review cadence.
Output Format
Strategic Context
One paragraph restating the company's current strategic priorities and how this roadmap serves them. Include key assumptions about customer needs.
Outcome-Focused Roadmap
Now (0-6 weeks) -- Committed
| Initiative | Outcome Statement | Key Metric | Owner | Dependencies |
|---|---|---|---|---|
| Original feature/project | Enable [segment] to [outcome] so that [impact] | Metric + target | Team | List |
Next (6-12 weeks) -- High Confidence
Same table format. Items here are planned but may shift based on learnings from Now.
Later (12+ weeks) -- Exploratory
Same table format. Items here are directional. Flexible time windows, no owners assigned yet.
Not Doing (and Why)
Initiatives explicitly excluded with reasoning. This section is as important as the roadmap itself.
Dependency Map
Visual or textual representation of which initiatives block or enable others.
Capacity Check
Summary of estimated effort vs. available capacity per time horizon.
Key Assumptions and Risks
Numbered list of what must be true for this roadmap to succeed.
Frameworks & Best Practices
- Outcomes over outputs. "Build advanced search filters" is an output. "Enable customers to find products 50% faster through intuitive discovery" is an outcome. Output-focused roadmaps create false precision and misalign teams around features rather than results.
- The outcome statement format. "Enable [customer segment] to [desired customer outcome] so that [business impact]." This forces clarity on who benefits, what changes, and why it matters.
- The "So What?" test. If you cannot articulate the outcome a feature drives, keep asking "So what?" until you reach real customer or business value. Multiple outputs may achieve one outcome -- focus on the outcome.
- RICE scoring.
- Reach: How many users/accounts will this affect per quarter? - Impact: On a scale of 0.25 (minimal) to 3 (massive), how much will this move the target metric? - Confidence: As a percentage, how sure are you about reach and impact estimates? - Effort: Person-weeks required. - Score: (Reach x Impact x Confidence) / Effort.
- Three-horizon structure. Now/Next/Later avoids the false precision of Gantt charts. Commit to Now, plan for Next, explore Later.
- Outcome roadmaps are resilient to change. Because they describe the "why" rather than the "what," they survive strategy pivots and new information better than feature roadmaps.
- Say no explicitly. A roadmap is defined as much by what it excludes. Maintain a "Not Doing" list and share the reasoning.
- Align with OKRs. Every outcome statement should map to a company or product OKR. Orphan initiatives signal misalignment.
- Revisit quarterly. Roadmaps older than one quarter are stale. Build in review cadences.
- Stage-appropriate planning.
- Pre-PMF: Roadmaps should be 4-6 week sprints focused on learning. No 12-month plans. - Post-PMF: Quarterly roadmaps tied to OKRs. Balance growth features with infrastructure and debt. - Scaling: Semi-annual roadmaps with cross-team coordination. Dependencies become the hard problem.
- Avoid roadmap theater. Do not create elaborate roadmaps to appease stakeholders if the team lacks the capacity or conviction to execute.
Related Skills
prd-writing-- Once an initiative is committed in the Now horizon, write a PRD to spec it out.mvp-scoping-- Before placing an initiative on the roadmap, scope its MVP to get a realistic effort estimate.user-research-synthesis-- Use customer insights to validate which outcomes matter most to your target segments.
Examples
Example 1: Output-to-outcome transformation
User: "Our Q2 plan says: build advanced search filters, implement AI recommendations, redesign dashboard. Help me turn this into an outcome roadmap."
Good output excerpt:
Transformation: - Output: "Build advanced search filters" --> Outcome: "Enable customers to find products 50% faster through intuitive discovery" - Output: "Implement AI recommendations" --> Outcome: "Increase average order value by 20% through personalized discovery" - Output: "Redesign dashboard" --> Outcome: "Help operators monitor all systems with 80% reduction in dashboard load time" Each outcome is now testable, measurable, and opens the door to alternative solutions beyond the originally specified feature.
Example 2: Feature backlog to roadmap
User: "Here's our feature backlog: Slack integration, SSO, reporting dashboard, mobile app, API v2, onboarding wizard. Help me build a roadmap."
Good output excerpt:
Now (0-6 weeks): | Onboarding Wizard | Enable new signups to reach first value within 10 minutes so that activation rate increases from 35% to 55% | Activation rate | Growth team | None | Not Doing This Quarter: - Mobile app -- Current usage data shows <5% of sessions are mobile. Revisit when mobile demand signal is stronger. Asking "So what?" reveals no compelling customer outcome today.