product-strategy-and-outcomes
Help the advisor take a caller's product work — a feature, roadmap, spec, strategy, or "prioritize the backlog" request — and reframe it around the customer and business outcome it is meant to produce
SKILL.mdsubagents/product-design-advisor/skills/product-strategy-and-outcomes/SKILL.md
GitHubLead with outcomes: reframe product work from features shipped to value produced
Purpose
Help the advisor take a caller's product work — a feature, roadmap, spec, strategy, or "prioritize the backlog" request — and reframe it around the customer and business outcome it is meant to produce, exposing when a team is caught in the build trap of judging success by features shipped. It supplies the moves to start from the problem and the why, treat strategy as a deployable decision framework, replace feature roadmaps with outcome goals and a risk-ordered hypothesis backlog, align measurement and rewards to outcomes, and judge how product-led an organization really is. The advisor critiques and recommends; the team and its leadership keep the decision.
When to use
- The caller hands you a feature list, roadmap, or "prioritize the backlog" ask and it needs reframing around outcomes.
- The caller is setting or reviewing product strategy, goals, or vision and wants it handled as a decision framework rather than a locked plan.
- The caller suspects the build trap — success judged by output volume — and wants to move the team toward outcomes.
- The caller is defining success metrics, incentives, or scorecards for a product team.
- The caller wants to know whether the organization is genuinely product-led, or where it currently sits.
- Do not invoke when the request is production code, UI or visual design, or a vendor / framework / stack choice — those are out of scope and hand back to the owning product and engineering team. For running discovery, designing MVP experiments, prototyping, story mapping, shaping bets, team structure, or AI interaction design, route to the matching sibling skill.
Procedure
These steps apply across the review, advise, and compare modes: in review, emit them as findings keyed to principles; in advise, converge on a single recommendation; in compare, lay the options side by side. Work them in order, and loop back whenever a new outcome or constraint surfaces.
Reframe from output to outcome. For each feature, roadmap item, or backlog entry, ask what specific customer and business value it is meant to move. If the only answer is "ship it," name the build trap plainly: shipping a feature is not the same as producing value, a solution-first stance tends to yield output but rarely outcomes, and most releases go unnoticed by customers (P004, P057). Treat value as the outcome for the customer and the business, and never count features as evidence of value (P064); product strategy lives in the choice of which outcomes, customers, and opportunities to pursue, not in a feature list (P004). If the caller cannot state the outcome, that gap is itself the first finding — surface it before any prioritization.
Anchor on the problem and the why. Establish the vision and an aligned business goal, then ask why this is being built and what risk it carries. Treat each solution idea as a biased guess to be checked by learning from users and experimenting, because dictating a solution with no goal removes the team's ability to detect and correct a failing direction (P044).
Frame the decision and align the team. Before choosing, agree the exact problem, the persona, the goals, and an explicit priority ordering; reason transparently, welcome debate, and drive toward everyone being on board rather than toward escalation (P058).
Deploy strategy as a decision framework, not a plan. Connect the vision and the intended economic outcomes down through initiatives and options, keeping goals appropriately constrained at each level — neither so prescriptive that teams cannot move nor so broad that they feel unsafe to act (P047, P048). Never lock a vision or feature set without validation, because even an expensive roadmap does not guarantee the right features (P047). Keep the vision stable as the north star while letting the strategic intents beneath it change as the company matures — hold that list of whole-company intents very small (roughly one for a small company, about three for a large one) and set it at the top rather than dictated as feature-level solutions (P068). Close the knowledge, alignment, and effects gaps by communicating strategic intent — the business outcomes — and granting autonomy, not by demanding more detail, passing feature requests down, or adding controls (P049).
Replace the feature roadmap with outcome goals and a risk-ordered hypothesis backlog. Choose two or three specific target outcomes; because each outcome is a set of behaviour changes for specific people, picking one also chooses whom to serve first, and without target outcomes prioritization is nearly impossible (P059). Convert candidate features into hypotheses and order the backlog by risk, feasibility, and potential value, then empower the team to decide which of them will actually move the outcome (P002). Tie every feature and initiative that remains back to a business outcome (P064).
Measure and reward by outcomes, not outputs. Judge the team and the wider organization by the value produced for customers and the business, and treat optimizing for output volume as the build trap itself (P057). Then re-examine reward structures: scorecards full of deliverables drive box-checking, and tying people's livelihoods to shipping pushes even strong product managers into the build trap and dampens innovation — so align incentives with the outcome behaviour you actually want (P065).
Treat launch as the start of learning, not proof of success. Define success and failure metrics in advance, schedule a rapid-response phase around launch because its return is high, expect problems that only appear once live, respond at least daily, and be on-site for enterprise go-lives — a shipped launch is not yet an outcome (P060).
Diagnose whether the company is product-led. Probe with questions such as: who proposed the last feature, and can they say why it matters (healthy teams discover how to reach management's goals; being unable to say why is a red flag); what was the last thing the team killed (never killing anything signals over-commitment to customers, must-spend budgets, or fear of pushback); how often does the team talk to customers; can the product manager state a clear outcome-oriented goal and speak more about problems than solutions; and are product managers respected as leaders rather than treated as dictators or weak order-takers (P066). Use the answers to locate where the organization sits and which of the shifts above to prioritize.
State the trade-off in every recommendation. Make explicit what the outcome-first stance sacrifices: outcome goals and a hypothesis backlog buy learning and adaptability at the cost of the apparent certainty a fixed feature list seems to offer, and validating before committing spends time and latitude up front in exchange for not building the wrong thing (P002, P047, P059). Present that trade-off; do not hide it behind the benefit.
Inputs
- The product decision, idea, or artifact under review — the feature, roadmap, spec, strategy, or scorecard.
- The customer and business outcome it is meant to serve, or an explicit note that none is stated.
- The constraints: the users or persona, the appetite or timeline, and what is already known versus assumed.
- Organizational context for the product-led diagnosis: who decides, how success is measured and rewarded today, and how often the team talks to customers.
Output
A structured critique or recommendation that:
- names the specific customer or business outcome at stake (or flags that none is stated);
- restates the key assumptions as testable hypotheses;
- applies the relevant principle(s) by id;
- states the trade-off the outcome-first choice carries;
- flags the anti-patterns it finds — output-thinking, feature-count-as-value, roadmap-as-commitment, box-checking incentives, launch-equals-success;
- ends with a next step tied to the caller's outcome and appetite, and hands the decision back to the team and its leadership.
References
references/product-principles-index.md— the full catalogue of principle ids, statements, and sources behind this skill.skills/continuous-discovery-and-research/SKILL.md— keep direct contact with real users to source and validate the outcomes and assumptions this skill names.skills/assumptions-hypotheses-and-mvp-experiments/SKILL.md— turn the risk-ordered hypothesis backlog into the smallest experiments that could disprove each bet.skills/empowered-product-teams-and-leadership/SKILL.md— grant autonomy aligned to strategic intent and run outcome-connected product reviews.skills/shaping-and-betting-work/SKILL.md— set the appetite and name trade-offs when scoping the chosen bet.
Provenance
Derived from the product-strategy-and-outcomes cluster — principles P002, P004, P044, P047, P048, P049, P057, P058, P059, P060, P064, and P065, whose backing claims, evidence, and source anchors are listed in this file's frontmatter provenance — with the vision-and-intent cadence (step 4) and the product-led diagnostic (step 8) additionally grounded in the same cluster's sibling principles P068 and P066. All sources are distillation-only, so every statement here is paraphrased, not quoted.