Kill, bang, marry is a prioritization framework that helps people choose among options by classifying each as one of three categories: kill (stop or avoid), bang (actively pursue now), or marry (commit to long term while balancing tradeoffs). Used widely in product strategy, career planning, budgeting, and personal decision-making, the approach emphasizes explicit tradeoffs, clarity over consensus, and ongoing reassessment rather than a single perfect formula.
Defining Kill, Bang, Marry
At the highest level, the framework assigns each option to a role that signals the recommended level of investment and attention. The labels are intentionally blunt so that discussions stay action oriented.
Kill
Options to kill should be actively stopped, avoided, or de‑prioritized. This includes initiatives that no longer align with strategy, consume disproportionate resources for limited return, or block more valuable work. Labeling something as kill is not necessarily permanent—some experiments may be killed now and revisited later under different conditions—but it signals an immediate, evidence based pause or stop.
Bang
Bang items are worth doing now. These are high priority opportunities, fixes, or bets where the expected value justifies near term investment of time, budget, or capacity. A bang option should have a clear hypothesis, defined success metrics, and an estimated horizon for meaningful impact or learning.
Marry
Marry choices are long term commitments that typically carry higher costs and dependencies, often involving sustained investment over quarters or years. Examples include platform migrations, major vendor relationships, or personal commitments such as partnerships or relocation. Marry decisions usually require alignment across teams, processes, and budgets, and they change the default path for an extended period.
When to Use the Framework
The framework works best when you need a simple, fast way to align on relative priorities among several options. It surfaces tradeoffs explicitly, making it useful in product roadmaps, portfolio reviews, hiring choices, budgeting, and personal decisions like feature sets, projects, or life choices.
Product and Roadmap Context
In product management, kill bang marry can help balance experimentation (bang), maintenance or sunsetting work (kill), and strategic bets (marry). By classifying initiatives, teams can communicate capacity constraints, clarify owners, and define gates for reevaluation without conflating short term experiments with long term commitments.
Personal and Career Decisions
Professionals use the framework to prioritize roles, projects, or skills. A job opportunity might be bang if it accelerates growth with manageable risk; a long term career path or relocation can be marry; and tasks or responsibilities that no longer contribute may be kill. This structure can reduce decision noise while preserving flexibility.
How to Apply It Step by Step
Applying the framework effectively depends on clear criteria, shared context, and a bias toward revisiting decisions as new evidence arrives.
- Define the decision scope and time horizon: clarify whether this is a quarterly roadmap, an annual portfolio review, or a personal choice with multiyear implications.
- List options and constraints: enumerate candidates, resource limits, dependencies, and non‑negotiable requirements.
- Set evaluation criteria: choose and communicate a small set of measures such as impact, cost, risk, time to value, strategic alignment, and optionality.
- Classify options: assign each option to kill, bang, or marry based on the criteria, noting key assumptions.
- Define next actions: for kill items, specify what to stop or avoid; for bang items, define experiments and milestones; for marry items, outline governance, timelines, and review checkpoints.
- Set review cadence: schedule periodic reassessments so classifications can change as conditions evolve.
Pros and Cons at a Glance
The framework is valued for speed and clarity but has limitations that are important to manage.
| Aspect | What It Offers | Practical Limitations |
|---|---|---|
| Decision Clarity | Simple labels reduce ambiguity and alignment time | Labels can oversimplify nuanced options |
| Speed | Fast to apply in workshops and written exercises | Rushing can omit critical context or data |
| Transparency | Makes tradeoffs and assumptions explicit | Assumptions must be documented to avoid false clarity |
| Flexibility | Supports reclassification as new evidence appears | Without reviews, classifications can become stale |
| Coordination | Helps teams communicate capacity and priorities | Requires shared definitions and agreement on criteria |
Common Pitfalls and Mitigations
Even a simple framework can mislead if used without rigor or empathy.
- Labeling as kill too quickly: It can shut down learning. Mitigate by distinguishing between kill now, pause for reconsideration, and sunset with a clear timeline.
- Calling everything bang: Overcommitment leads to fatigue and risk concentration. Use caps on concurrent bang items and reserve capacity for maintenance and learning.
- Marry without exit criteria: Long commitments can become sunk‑cost traps. Define review gates, success thresholds, and contingency plans before marrying.
- Ignoring optionality: Overcommitting to a single path reduces flexibility. Preserve small, low cost experiments even when you marry a direction.
- Lacking shared criteria: Different people using different rules causes confusion. Agree on evaluation metrics and how to weight them before classifying options.
Practical Tips for Better Decisions
- Pair the framework with a lightweight scoring rubric for bang and marry items to compare options quantitatively while retaining the simple labels.
- Document assumptions for each classification so that reviews focus on evidence rather than opinion.
- Use timeboxing for kill and bang work to preserve capacity for strategic marry initiatives.
- Balance your portfolio: aim for a healthy mix of kill (removing drag), bang (learning and short term value), and marry (building durable advantage) based on capacity and risk tolerance.
- Review regularly: schedule quarterly or biannual portfolio reviews to reclassify items as conditions change.
Examples in Context
Examples help illustrate how classifications can differ by context and how they evolve over time.
Product Roadmap Example
A product team might classify a minor UI improvement as bang (ship within one sprint), a new analytics integration as marry (multi quarter project with cross‑team dependencies), and an underused legacy feature as kill (sunset after deprecation communication). As user data arrives, the team can reclassify items during scheduled reviews.
Hiring and Team Planning Example
For a growing team, a critical role needed immediately to unblock work might be bang, a long term people strategy and succession plan could be marry, and low impact recurring reports might be kill or delegated to streamline capacity.
Personal Finance Example
An individual might classify paying down high interest debt as bang, funding retirement savings as marry, and a subscription that no longer delivers value as kill. This makes tradeoffs visible and supports more intentional allocation of limited resources.
When the Framework Should Not Be the Default
Kill, bang, marry is most effective as a lightweight alignment tool rather than a rigorous decision model. For complex, data intensive, or high risk investments, complement it with deeper analysis, financial modeling, or structured decision frameworks. It is a compass, not a detailed map, and should be used in conjunction with domain expertise and quantitative evidence.
Summary and Takeaways
The kill, bang, marry framework is an evergreen tool for clarifying priorities by forcing explicit choices among stop, now, and long term commitments. It favors speed and transparency while surfacing tradeoffs, but it works best when classifications are documented, assumptions are stated, and decisions are revisited regularly. Used thoughtfully, it supports better product roadmaps, more coherent strategies, and more intentional personal choices.