Product Reference
Product Management Frameworks
A product framework is a structured argument, not an oracle: RICE and MoSCoW rank work, Jobs to be Done and Opportunity Solution Trees frame the problem, and OKRs and North Star metrics define success. Each has a characteristic failure mode, which is what decides whether it helps or produces a defensible-looking ranking that is simply wrong.
Browse by Category
All Product Management Frameworks (28)
Prioritisation (9)
Deciding what to build next, and what not to.
| Framework | Kind | What it does | Watch out for |
|---|---|---|---|
RICE | Scoring | Reach × Impact × Confidence ÷ Effort. Forces you to state each factor rather than argue by conviction. | False precision — the inputs are estimates |
ICE | Scoring | Impact, Confidence, Ease. RICE without reach — faster, cruder, fine for a small backlog. | Everything scores 7 or 8 |
MoSCoW | Categorical | Must, Should, Could, Won't. Best used to agree what is out of scope, which is its real value. | Everything becomes a Must |
Kano model | Categorical | Splits features into basics, performance and delighters. Explains why fixing basics wins no praise. | Delighters become basics over time |
Value vs Effort | Matrix | A 2×2 of impact against cost. The quickest way to make a room agree on the obvious wins. | Everyone plots their own item top-left |
Weighted scoring | Scoring | Score against weighted criteria you choose. Flexible, and only as good as the weights. | Weights get tuned to justify a decision |
Cost of Delay / WSJF | Economic | What does waiting cost per week? Sharp for sequencing when items are genuinely time-sensitive. | Hard to estimate honestly |
Now / Next / Later | Horizon | Three buckets instead of dates. Communicates sequence and confidence without promising a calendar. | "Later" becomes a graveyard |
Eisenhower matrix | Matrix | Urgent against important. A personal time-management tool that also works on a backlog. | Urgent crowds out important |
Kind
Scoring
What it does
Reach × Impact × Confidence ÷ Effort. Forces you to state each factor rather than argue by conviction.
Watch out for
False precision — the inputs are estimates
Kind
Scoring
What it does
Impact, Confidence, Ease. RICE without reach — faster, cruder, fine for a small backlog.
Watch out for
Everything scores 7 or 8
Kind
Categorical
What it does
Must, Should, Could, Won't. Best used to agree what is out of scope, which is its real value.
Watch out for
Everything becomes a Must
Kind
Categorical
What it does
Splits features into basics, performance and delighters. Explains why fixing basics wins no praise.
Watch out for
Delighters become basics over time
Kind
Matrix
What it does
A 2×2 of impact against cost. The quickest way to make a room agree on the obvious wins.
Watch out for
Everyone plots their own item top-left
Kind
Scoring
What it does
Score against weighted criteria you choose. Flexible, and only as good as the weights.
Watch out for
Weights get tuned to justify a decision
Kind
Economic
What it does
What does waiting cost per week? Sharp for sequencing when items are genuinely time-sensitive.
Watch out for
Hard to estimate honestly
Kind
Horizon
What it does
Three buckets instead of dates. Communicates sequence and confidence without promising a calendar.
Watch out for
"Later" becomes a graveyard
Kind
Matrix
What it does
Urgent against important. A personal time-management tool that also works on a backlog.
Watch out for
Urgent crowds out important
Discovery & Framing (7)
Finding the right problem before committing a team.
| Framework | Kind | What it does | Watch out for |
|---|---|---|---|
Jobs to be Done | Framing | People "hire" a product for a job. Reframes the competitor set around the job, not the category. | Retrofitted to justify a plan |
Opportunity Solution Tree | Mapping | Connects outcome to opportunities to solutions to experiments, so a feature traces back to a goal. | Grows faster than it is pruned |
Double Diamond | Process | Diverge then converge, twice — once on the problem, once on the solution. Stops solution-first thinking. | Teams skip the first diamond |
Riskiest Assumption Test | Validation | Test the belief that would sink the idea, before building anything. Usually cheaper than an MVP. | The riskiest assumption is misidentified |
Working Backwards (PR-FAQ) | Framing | Write the press release and FAQ first. If it is not compelling on paper, it will not be in code. | Becomes a writing exercise |
Impact mapping | Mapping | Goal → actor → impact → deliverable. Makes the assumption chain behind a feature explicit. | Workshop output nobody revisits |
Opportunity sizing | Estimation | How many people, how often, how painful. Turns a good idea into a ranked one. | Top-down market numbers mean little |
Kind
Framing
What it does
People "hire" a product for a job. Reframes the competitor set around the job, not the category.
Watch out for
Retrofitted to justify a plan
Kind
Mapping
What it does
Connects outcome to opportunities to solutions to experiments, so a feature traces back to a goal.
Watch out for
Grows faster than it is pruned
Kind
Process
What it does
Diverge then converge, twice — once on the problem, once on the solution. Stops solution-first thinking.
Watch out for
Teams skip the first diamond
Kind
Validation
What it does
Test the belief that would sink the idea, before building anything. Usually cheaper than an MVP.
Watch out for
The riskiest assumption is misidentified
Kind
Framing
What it does
Write the press release and FAQ first. If it is not compelling on paper, it will not be in code.
Watch out for
Becomes a writing exercise
Kind
Mapping
What it does
Goal → actor → impact → deliverable. Makes the assumption chain behind a feature explicit.
Watch out for
Workshop output nobody revisits
Kind
Estimation
What it does
How many people, how often, how painful. Turns a good idea into a ranked one.
Watch out for
Top-down market numbers mean little
Delivery (6)
Turning a decision into something a team can build.
| Framework | Kind | What it does | Watch out for |
|---|---|---|---|
User story | Requirement | As a… I want… so that… A placeholder for a conversation, not a specification. | Written as a spec, never discussed |
Acceptance criteria | Requirement | The conditions that make a story done, including the edge cases. Where ambiguity actually gets removed. | Happy path only |
Story mapping | Planning | Lays stories along the user journey, so a release slice is coherent rather than a list of parts. | Done once, then abandoned |
Definition of Ready / Done | Agreement | Shared entry and exit criteria for work. Prevents the "is it finished?" argument recurring weekly. | A gate rather than an agreement |
MVP | Scoping | The smallest thing that tests the core assumption. Minimum AND viable — most teams drop the second word. | Means "version one, but rushed" |
RACI | Roles | Responsible, Accountable, Consulted, Informed. Useful when it is genuinely unclear who decides. | Everyone is Consulted |
Kind
Requirement
What it does
As a… I want… so that… A placeholder for a conversation, not a specification.
Watch out for
Written as a spec, never discussed
Kind
Requirement
What it does
The conditions that make a story done, including the edge cases. Where ambiguity actually gets removed.
Watch out for
Happy path only
Kind
Planning
What it does
Lays stories along the user journey, so a release slice is coherent rather than a list of parts.
Watch out for
Done once, then abandoned
Kind
Agreement
What it does
Shared entry and exit criteria for work. Prevents the "is it finished?" argument recurring weekly.
Watch out for
A gate rather than an agreement
Kind
Scoping
What it does
The smallest thing that tests the core assumption. Minimum AND viable — most teams drop the second word.
Watch out for
Means "version one, but rushed"
Kind
Roles
What it does
Responsible, Accountable, Consulted, Informed. Useful when it is genuinely unclear who decides.
Watch out for
Everyone is Consulted
Metrics & Goals (6)
Defining success, and finding out whether it happened.
| Framework | Kind | What it does | Watch out for |
|---|---|---|---|
North Star metric | Metrics | One measure that captures delivered value. Aligns a team faster than any roadmap. | Chosen because it is easy to move |
AARRR (pirate metrics) | Metrics | Acquisition, Activation, Retention, Referral, Revenue. A funnel view of the whole business. | Retention gets least attention |
HEART | Metrics | Happiness, Engagement, Adoption, Retention, Task success. Google's UX-centred counterpart to AARRR. | Measured without a target |
OKRs | Goal setting | An objective with measurable key results. Works when outcome-based, fails when it lists features. | Key results that are just deliverables |
Cohort analysis | Metrics | Groups users by when they joined and tracks them over time. Separates growth from retention. | Read as one aggregate line |
Product-market fit survey | Signal | How disappointed would you be if this vanished? Above ~40% "very" is the common rule of thumb. | Run on the wrong segment |
Kind
Metrics
What it does
One measure that captures delivered value. Aligns a team faster than any roadmap.
Watch out for
Chosen because it is easy to move
Kind
Metrics
What it does
Acquisition, Activation, Retention, Referral, Revenue. A funnel view of the whole business.
Watch out for
Retention gets least attention
Kind
Metrics
What it does
Happiness, Engagement, Adoption, Retention, Task success. Google's UX-centred counterpart to AARRR.
Watch out for
Measured without a target
Kind
Goal setting
What it does
An objective with measurable key results. Works when outcome-based, fails when it lists features.
Watch out for
Key results that are just deliverables
Kind
Metrics
What it does
Groups users by when they joined and tracks them over time. Separates growth from retention.
Watch out for
Read as one aggregate line
Kind
Signal
What it does
How disappointed would you be if this vanished? Above ~40% "very" is the common rule of thumb.
Watch out for
Run on the wrong segment
Frequently Asked Questions
Which prioritisation framework is best?
None of them decides for you. RICE is the most defensible when you can estimate reach, Value vs Effort is fastest for getting a room to agree, and MoSCoW is best at agreeing what is out of scope. Pick one, use it consistently, and own the call it suggests rather than hiding behind the score.
Are these frameworks actually useful?
As a way to make reasoning explicit, yes. Their value is that they force you to state reach, impact, confidence or cost out loud, where someone can disagree with the input. Their danger is false precision — a number derived from four guesses is still four guesses.
What is the difference between OKRs and KPIs?
A KPI is an ongoing health measure you watch continuously; an OKR is a time-boxed ambition with an objective and measurable key results. Confusing them produces OKRs that are just a list of metrics you were already tracking, which changes nobody's behaviour.
What does MVP actually mean?
The smallest thing that tests your riskiest assumption — minimum and viable. In practice it usually degrades to "version one, but rushed", which is a different and much worse idea, because a rushed version one tests nothing and damages trust.