Resources

Quick reference

The tables worth pinning above your desk.

Prioritisation scoring, side by side

Model	Formula	Use it when	Its blind spot
RICE	(Reach × Impact × Confidence) ÷ Effort	Mixed backlog, many request types	Favours broad and shallow work
ICE	Impact × Confidence × Ease	Fast triage, growth experiments	Every input is an opinion
WSJF	Cost of Delay ÷ Job Size	Deadline-driven or compounding work	False precision from stacked estimates
Value / Effort	Two axes, four quadrants	Workshops, quick alignment	"Value" means something different to everyone

Full breakdown in prioritisation frameworks and where each one lies to you.

Metric definitions worth agreeing on

Metric	Usually means	Ask before trusting it
Activation	User reached first real value	What counts as value, and who decided?
Retention	Still active after N periods	Cohort or snapshot? Snapshots hide leaks
DAU / MAU	Engagement intensity ratio	Does daily use actually make sense for this product?
Churn	Customers lost in a period	By count or by revenue? They tell different stories
NPS	Willingness to recommend	Who responded — and who didn't?
Feature adoption	Share of users who used it	Used once, or came back? Only the second counts

Write down what each metric excludes — internal accounts, trials, bots. Six months on nobody remembers, and someone will decide on a number that never meant what they think.

Scrum event timeboxes

Event	One-month sprint	Two-week sprint	Your job in it
Sprint Planning	8 hours	4 hours	Bring the goal and the why
Daily Scrum	15 min	15 min	Listen; unblock decisions
Sprint Review	4 hours	2 hours	Bring stakeholders, change the backlog
Retrospective	3 hours	90 min	Attend as a peer, not a manager
Refinement	Not an event — roughly 10% of capacity		Your highest-leverage hour

Steal these templates

Copy the text straight out. No download, no email required.

Discovery interview opener

"Tell me about the last time you [did the job]." Then: what happened just before? What did you try first? What did you do afterwards? Never ask whether they would use a thing you are about to describe.

The full question set →

Acceptance criteria skeleton

Given [starting state], when [action], then [observable outcome]. One scenario for the happy path, one for the likeliest error, one for the boundary.

See worked examples →

Opportunity, not solution

An opportunity is phrased in the customer's words and came from a real conversation: "I have to check three places before I can answer my manager." If it names a feature, walk it back up until you find the need underneath.

Building the tree →

Roadmap item card

Outcome · horizon (Now / Next / Later) · confidence (high / medium / low) · the metric that should move. Confidence on the card does more for stakeholder trust than any format change.

Roadmap formats →

Job statement

When [situation], I want to [motivation], so I can [expected outcome]. If the "when" is generic, you have not talked to enough people about specific incidents yet.

Getting it right →

Retrospective agenda (60 min)

5 min set the stage · 15 min gather data · 15 min generate insight · 15 min decide what to do · 10 min close. One action, one owner, one sprint.

Seven formats to fill it →

Worth reading

The short list, if you only get through a few.

Continuous Discovery Habits, Teresa Torres — the clearest treatment of opportunity solution trees and weekly customer contact.
Inspired and Empowered, Marty Cagan — how strong product organisations are actually structured, and why most are not.
The Scrum Guide — nineteen pages, free, and the definitive answer to most "is this still Scrum?" arguments. At scrumguides.org.
The Agile Manifesto and its twelve principles — a single page from 2001 that still settles more arguments than most frameworks do.
Escaping the Build Trap, Melissa Perri — on the difference between shipping features and delivering outcomes.

Templates, checklists and primary sources