Vision strategy

Product Vision & Strategy Skill

How this skill behaves (read first)

This is a generative skill, and vision/strategy is where an AI assistant produces its most confident, least useful output. Two default failure modes dominate:

1. The inspirational poster. Asked for a vision, Claude writes a buzzword sentence — "be the world's most loved, best-in-class platform that delights customers with seamless experiences" — that sounds good and guides zero decisions. If any feature can be justified by it, it's too vague to be a vision.
2. Conflation and feature-listing. Claude treats vision, strategy, roadmap, and mission as interchangeable, and writes "strategy" as a list of features to build rather than a set of deliberate choices about where to play, how to win, and what not to do. Strategy that targets "everyone" and skips trade-offs isn't strategy.

A real vision is a specific, stable, user-centric destination; a real strategy is a focused set of interconnected, trade-off-driven choices grounded in evidence and tied to outcomes. So this skill gates:

1. Establish context — vision, strategy, or both; company shape; product stage; market type; altitude.
2. Apply the always-true core — the artifact distinctions, the laddering, the focus-and-trade-off discipline, the stability rule.
3. Surface the context-dependent decisions (which framework, which communication format, portfolio/B2B/stage/depth) with trade-offs, and let the user choose. Running every framework at once is the failure mode.

Then it hands off to pm-okr-metric-validity-audit (the success metrics it names), pm-prioritization-rigor-audit (the themes/focus areas it sets, which drive what gets built), and pm-assumption-rigor-audit (the market-demand and competitive bets the strategy rests on).

Scope: this skill owns the vision artifact and the strategy artifact, and the choices inside them. It defers the competitive/market analysis that feeds strategy to pm-competitive-analysis, the research/insight-gathering to pm-discovery, problem framing to pm-problem-statement, the prioritization decision itself to pm-prioritization, the OKR/KPI artifacts to pm-okrs-kpis, the roadmap artifact and sequencing to pm-roadmap, pricing to ux-pricing, and go-to-market to pm-gtm-plan.

---

Step 0 — Establish context before writing

Ask if not known; state the assumption if proceeding without an answer:

• Vision, strategy, or both? They're different artifacts at different altitudes (see the core below). Writing one when the user needs the other — or blending them into mush — is the first mistake. If they ask for "a strategy" but have no vision, flag that the vision is the anchor strategy derives from.
• What's the company shape — single-product or multi-product? A single-product company's product vision nearly equals its company vision; a multi-product portfolio needs each product's vision/strategy to ladder up to the company vision without colliding with siblings.
• New product or existing/pivot? New products derive vision from market research, problem interviews, and a small founder/early-team group; existing products have real usage data, feedback patterns, and a broader stakeholder set to draw on.
• B2B or B2C? B2B strategy serves multiple stakeholders (end user, champion, economic buyer), longer cycles, higher switching costs; B2C weights individual decision-makers, viral acquisition, and network effects.
• What altitude / scope? A feature-area vision shouldn't span the whole company; a company-wide product vision can be broad. Scope determines who must align and which decisions the vision governs.

---

The always-apply core (true for any vision or strategy)

Keep the four artifacts distinct (and ladder them)

• Don't conflate vision, strategy, roadmap, mission. They form a hierarchy at decreasing altitude and increasing change-frequency: company vision → company strategy → product vision → product strategy → roadmap/features. Vision is the aspirational destination (stable for years). Strategy is how you'll get there (the choices; evolves every few quarters). Roadmap is what and when (changes often). Mission is why you exist today; vision is the future you're creating. When vision is absent, roadmaps become disconnected feature lists; when strategy masquerades as vision, you lose the inspirational anchor.
  • ❌ "Our vision: ship AI search, a mobile app, and SSO by Q4." (That's a roadmap.)
  • ✅ "Our vision: every analyst finds the answer in their data in seconds, without writing a query." (A destination; many strategies could reach it.)
• Derive product vision from company vision. Identify which slice of the company vision your product makes tangible; it should read as a natural, more-specific extension. Translate company strategy (markets, positioning, business model) into product choices (which problems, which users, which capabilities first).

Make the vision specific, user-centric, and memorable

• Three traits: user-centric, simple, future-facing. Describe the changed user future, not your feature set, in language anyone can recall and repeat without the doc.
  • ❌ "Use integrated data systems to drive optimized wellness outcomes."
  • ✅ "Help people track their health in one simple app." / LinkedIn: "Create economic opportunity for every member of the global workforce."
• Ban empty superlatives. "Best-in-class," "world-leading," "seamless," "delight" sound impressive and mean nothing. Use concrete user outcomes instead.
  • ❌ "Empower teams to collaborate." ✅ "Help remote teams make decisions without meetings."
• The decision-filter test. A vision works only if it can reject a feature. If someone can justify any feature with it, it's too vague — tighten until a feature idea clearly aligns or conflicts.
• 3–5 year horizon. Shorter than ~3 years is a roadmap, not a vision.

Make the strategy a set of focused, trade-off-driven choices

• Strategy = interconnected choices, not a feature list. Name the winning aspiration (vision), where to play (segment + market), how to win (differentiation), the capabilities/resources required, growth mechanism (product-led vs. sales-led + channels), and key metrics — and check the elements reinforce each other.
• Pick ONE primary segment; "everyone" means no one. Define the target by the job/outcome they seek and their constraints, not just demographics. Win one segment and reach product-market fit before expanding (Instagram → iPhone users sharing photos fast, not "all mobile users").
• Choose a differentiation path and defend it. Cost leadership, unique value, or focus — not "better at everything." Back it with a barrier (proprietary tech, network effects, data, brand, partnerships). Apply the "can't or won't" test: competitors either can't copy you (real barrier) or won't (it conflicts with their model).
• State the trade-offs — what you deliberately won't do. This is the part Claude omits and the part that creates focus. "We won't serve enterprise," "we won't build native fidelity first," IKEA won't pre-assemble. No trade-offs = no strategy.
• Set 3–5 focus areas, max. More than that spreads teams thin and executes nothing well. Focus areas flow from the segment/problem/differentiation choices.
• Tie strategy to outcomes, and ground it in evidence. Frame goals as outcomes ("increase retention 10%") not outputs ("launch a loyalty program"), and connect each to a metric. Build it on real customer insight and competitive/market analysis, with assumptions flagged for testing — don't strategize in a vacuum (defer the analysis and research to the skills named above).

Set the stability rule

• Stubborn on vision, flexible on strategy. Vision should hold 3–5 years and survive bumps; strategy is reviewed (at least) quarterly and changes as you learn. Distinguish a vision pivot (the problem itself doesn't matter to customers — rare, serious: Burbn→Instagram) from a discovery pivot (a different path to the same destination — frequent, normal: change monetization, growth tactic, or feature focus).
• Define change-criteria up front so the call is evidence-based, not emotional: what validation signal would prove the problem doesn't matter (vision), and what metric/learning would prove the approach isn't working (strategy). Persisting too long wastes resources; pivoting too fast never lets a good idea prove itself — most teams give up too soon, not too late.

---

The context-dependent decisions (surface, don't auto-apply)

Present each with its trade-off and a recommendation tied to Step 0; let the user choose. Applying every framework and every communication format at once is the failure mode.

Decision	Apply when	Avoid / adapt when	Default recommendation
Which strategy framework	Playing to Win (5 cascading choices) for a full strategy; Porter's Five Forces for industry attractiveness/defense; Blue Ocean to create uncontested space; Kano to classify must-have/performance/delighter features; Three Horizons to balance now/next/future; Product Strategy Canvas to capture it all on one page	Running all of them as ritual — "framework theater" that buries the actual choices	Pick the 1–2 that answer the question; Product Strategy Canvas as the documenting container
Vision communication format	Written statement as the foundation; story/storyboard to make the user journey tangible; video for all-hands/investor emotional pitch; visiontype (idealized prototype) to show a 3–5-yr future when teams need a concrete north star	Using one format for every audience, or polishing a video before the statement is even clear	Start with the concise written statement, then adapt to the audience; reinforce the same core message across formats
Vision statement form	Template — "For [target] who [need], [product] is a [category] that [benefit]. Unlike [alternative], we [differentiation]." — when you want rigor and comparability	Forcing the template when a one-line narrative ("entertain the world") lands harder	Draft with the template to force completeness, then cut to the most memorable phrasing
Single vs. multi-product laddering	Single-product: vision ≈ company vision, keep them tight	Multi-product: each product needs its own vision/strategy that ladders up without overlapping siblings; watch for products competing for the same users	Make the product↔company link explicit either way; define clear product boundaries in a portfolio
B2B vs. B2C emphasis	B2B: multi-stakeholder value, champions, integration/security/scalability, longer validation	B2C: individual decision-maker, viral/network effects, fast iteration	Match the strategic emphasis to the buyer reality; don't apply a B2C playbook to an enterprise product
New vs. existing/pivot inputs	New: market reports, competitor scan, problem interviews, small group (6–8), workshop or week-long sprint	Existing/pivot: usage analytics, feedback patterns, broader cross-functional stakeholders	Match inputs and the group to the stage; keep the core vision group ≤8 for decisiveness
Depth — full doc vs. summary	Full strategy doc (segments, value prop, differentiation, focus, trade-offs, metrics, rough timeline) for alignment; 2–3-sentence summary for daily internalization	Producing a long document nobody reads, or a one-liner with no rationale	Write both: the short version teams remember + the detailed version with reasoning and assumptions

A few contexts flip several defaults at once: a multi-product B2B company needs portfolio laddering and multi-stakeholder strategy and a documented full strategy; a pre-PMF startup wants a tight written vision, one segment, one differentiation bet, and a lightweight strategy it expects to pivot.

---

Validate the result (orchestration)

Hand-offs name each lens by its installable skill name. Invoke one only if that skill is installed; if it isn't, this skill's own core already carries these rules — proceed without it rather than blocking.

After producing or revising the vision/strategy, hand it to the audit lenses rather than declaring it done. These are candidate lenses — posture per docs/orchestration-policy.md, or route the whole artifact through pm-product-review. Here assumption-rigor is the always-relevant lens (auto-runs) — strategy is bets; the other two are offered, tied to what the artifact actually contains. If the user invoked this skill for one specific thing, respect that scope.

• pm-assumption-rigor-audit (auto-runs) — strategy rests on bets about market demand, the target segment's problem, and the durability of the differentiation: are those load-bearing assumptions evidenced or asserted? Flag the riskiest for testing.
• pm-okr-metric-validity-audit (offer — if it names metrics) — the success metrics, KPIs, and North Star: valid outcome measures tied to the strategy, or vanity/output metrics? (Build the OKR/KPI artifacts themselves with pm-okrs-kpis.)
• pm-prioritization-rigor-audit (offer — if it sets themes/focus areas) — the 3–5 focus areas are prioritization choices that decide what gets built: evidence-based and traceable, or a feature-factory wish list / HiPPO pick? (Run the prioritization decision itself through pm-prioritization.)

Composition (per docs/orchestration-policy.md §9): this artifact sits high on the PM ladder and mostly feeds from upstream inputs and feeds into downstream artifacts rather than building them — the competitive analysis (pm-competitive-analysis), problem (pm-problem-statement), and discovery (pm-discovery) are upstream inputs; the roadmap (pm-roadmap), OKRs (pm-okrs-kpis), and pricing/GTM (ux-pricing, pm-gtm-plan) are downstream (offer to produce next, don't auto-generate). If the audits surface a conflict, resolve back toward the purpose: a vision exists to guide and reject decisions, and a strategy exists to make focused trade-offs — if the vision can justify anything and the strategy says no to nothing, neither is done.

---

Common do/don't patterns

❌ Don't	✅ Do
Write a buzzword poster ("best-in-class, seamless, world-leading")	Name a specific, user-centric, memorable destination that can reject a feature
Treat vision, strategy, roadmap, mission as the same thing	Keep the hierarchy distinct and ladder product → company
Write "strategy" as a list of features to build	Make deliberate choices: where to play, how to win, what you won't do
Target "everyone"	Pick one primary segment by job/outcome, win it, then expand
Claim you're better at everything	Choose one differentiation path with a defensible barrier ("can't or won't")
Omit the trade-offs	State explicitly what you deliberately won't do — that's the focus
List 10 focus areas	Hold to 3–5; more spreads teams thin
Run every strategy framework as ritual	Pick the 1–2 that answer the question
Use one vision format for all audiences	Start with the written statement, adapt to the audience
Change the vision every quarter	Stubborn on vision (3–5 yrs); flexible on strategy (quarterly); know vision-pivot vs. discovery-pivot
Ship vision/strategy unchecked	Hand off to okr-metric-validity + prioritization-rigor + assumption-rigor

---

Source lessons (Uxcel)

• What Is Product Vision?
• Defining the Product Vision
• Communicating the Product Vision
• Your Product Vision and Strategy
• Context of Product Vision and Product Strategy
• Aligning Teams with Product Vision & Strategy
• Lifetime of the Product Vision and Strategy
• How to Build a Product Strategy
• Building a Product Strategy
• Defining the Product Strategy