Product Management Playbooks
Practical, no-fluff playbooks built from real product workshops and real delivery. Each one runs 9-11 pages: a clear framework, a worked example, and a blank canvas to build your own.
Outcome-Based Roadmaps
Why customer behaviour, not shipped features, is what your roadmap should measure, and a practical method for building one.
Most roadmaps are lists of features and dates. They tell a team what to build and when, but say nothing about whether any of it will work. This playbook sets out a repeatable method for building a roadmap around outcomes instead: a structure that connects a product vision to day-to-day delivery, and keeps every feature honest about the problem it's meant to solve.
What's inside
- Why shipping a feature isn't success, and what the research says about how often features actually work
- The real difference between a roadmap and a release plan
- A 5-step method: vision → goals → outcomes → opportunities → validated solutions
- A fully worked example
- Common pitfalls that quietly turn an outcome-based roadmap back into a backlog
- A blank canvas to build your own
Beyond the Roadmap
A fair response to the "kill the roadmap" movement, from someone who has run one in a regulated, multi-stakeholder environment.
This is a companion to the first playbook in this series. It takes Teresa Torres's continuous discovery argument seriously, on its own terms, then makes the case for why its preconditions don't hold for regulated industries, enterprise B2B, or any environment with gated user access and shared decision-making, and sets out a hybrid model that keeps what actually works from both approaches.
What's inside
- A fair account of continuous discovery's real argument, not a strawman
- The preconditions the method requires, and why they rarely get discussed as loudly as the method itself
- Where those preconditions specifically break down in regulated and enterprise environments
- A hybrid model: an outcome-based roadmap as the compass, a realistic discovery cadence underneath it
- A worked comparison and common pitfalls
- A blank canvas to design your own hybrid cadence
Metrics That Matter
Which numbers actually tell you whether your product is working, and why the ones on your dashboard right now might not.
Most dashboards are full of numbers that go up and to the right and tell you almost nothing about whether your product actually works. This playbook sets out the metrics that matter, why they matter, and a framework for choosing which ones deserve your attention at your product's current stage.
What's inside
- Why no single metric tells the whole story, and how to choose the right ones for your stage
- The AARRR framework and the North Star Metric, and why they're not the same thing
- MAU/DAU and the "what counts as active" problem
- Conversion rate, and the difference between a discoverability problem and a desirability problem
- Churn (and why not all churn is bad)
- NPS and CSAT, properly defined and properly sourced
- A fully worked example and a blank canvas to build your own dashboard
Betting on Bayes
A prioritisation framework that scores your confidence honestly, so gut feel stops disguising itself as data.
Most roadmap priority calls come down to a gut feeling dressed up as a business case. This playbook sets out a Bayesian approach to prioritisation: a framework that forces you to show your working on confidence, rather than letting it hide behind a "high/medium/low" label.
What's inside
- Why cognitive bias makes prioritisation harder than it looks
- The Bayesian method, explained without the maths getting in the way
- The four inputs: Goal, Impact, Confidence, Effort, and how they relate to ICE and RICE scoring
- A fully worked example, including why the highest-confidence idea on the list still finishes last
- Common pitfalls
- A blank canvas, with an evidence-weighting exercise, to score your own initiatives
The Multiplier Effect
Product managers don't generate value on their own. They multiply the value created by the people around them, if they take the time to understand them.
The actual value in your product, the code, the design, the deals, the support, is created by the people around you, not by you. This playbook is a practical guide to building genuine empathy with those stakeholders, so you can make them more effective, and in turn become more effective yourself.
What's inside
- Why product managers are value multipliers, not value generators
- Who actually counts as a stakeholder (the list is longer than you think)
- Mendelow's Power-Interest grid, for deciding who deserves how much of your time
- Three ways to build empathy: informal interactions, formal research (with a full set of interview questions), and regular checkpoints
- A fully worked stakeholder empathy plan
- Common pitfalls
- A blank canvas to map your own stakeholders
The Craft of Product Management
Five practical skills that separate product managers who react to their product from those who actually shape it.
Most product manager job descriptions list responsibilities. Almost none of them list the actual skills the role demands, and they span a far wider range than people expect. This playbook breaks the role down into five distinct skills, where they come from, and how to build them deliberately.
What's inside
- Data Analysis, and the truth behind the "what gets measured gets managed" myth
- Product Immersion, and why "dogfooding" has been standard practice since 1988
- Keeping the Balance between present delivery and future planning
- Benchmarking properly, with context, instead of reinventing the wheel
- People Management without direct authority
- A worked example turning all five skills into weekly habits
- Common pitfalls and a blank canvas to build your own operating rhythm
Built in Four Weeks
What building the UK's furlough and Eat Out to Help Out platforms in weeks taught me about speed, and why most "we need six months" arguments don't hold up.
In spring 2020, HMRC needed to build a service capable of paying millions of people, at hyperscale, with no room for a phased rollout. It had about four weeks. I was a product owner on the UK Government team that delivered it.
This playbook isn't "move fast and skip the boring parts." It's closer to the opposite: speed at that scale isn't improvised in a crisis, it's bought years in advance, and spent deliberately once the crisis arrives. Three things made four-week delivery possible, and only one of them happened during the crisis itself.
What's inside
- Why the six-month excuse doesn't survive contact with a real crisis
- The three things that actually made four-week delivery possible: pre-built platform capability, compressed (not skipped) discovery, and disciplined temporary trade-offs
- A worked example comparing a normal research cadence to a crisis one, and what has to stay exactly the same between them
- Common pitfalls, including what happens when teams copy the speed without the capability behind it
- A crisis-readiness audit: three questions to ask your own organisation, plus a trade-off ledger to build your own
The Jagged Frontier
A grounded look at what AI actually changes about product management, and the research showing why blind trust in it can make you worse at your job, not better.
Most writing on AI and product management is hype or panic. This playbook is neither. It's built on a rigorous 2023 Harvard Business School and Boston Consulting Group study that tested 758 consultants with GPT-4 on real, complex tasks, and found something stranger and more useful than either extreme: AI's capability is uneven in ways that are genuinely hard to predict, and trusting it in the wrong place makes skilled people worse at their jobs, not better.
What's inside
- The research behind the "jagged technological frontier," and why fluent wrong answers are more dangerous than obviously bad ones
- A practical map of which PM tasks sit safely inside AI's capability, and which don't
- Two working principles for delegating pattern-work while keeping the judgement calls
- A worked example auditing a typical week of PM work
- Common pitfalls, including the one that actually predicts AI failure
- A blank canvas to audit your own week
Escape the Factory
A short, honest diagnostic for finding out whether your organisation ships features or creates outcomes, and what to do about the answer.
Most feature factories don't know they're feature factories. They're usually fast, disciplined, and genuinely agile in their ceremonies. What they're missing is any mechanism that would tell them if the wrong feature just shipped efficiently. This playbook is a nine-question diagnostic, built on Melissa Perri's "build trap" and Marty Cagan's empowered-teams work, for finding out where you actually stand.
What's inside
- Why the feature factory pattern rarely announces itself, and accumulates one reasonable-sounding request at a time
- A nine-question diagnostic across three areas: where ideas come from, how success is measured, and team autonomy
- How to read your score, and why a high score is a structural problem, not a character flaw
- A fully worked example scorecard
- Common pitfalls
- A blank canvas to score your own team
The Strangler Method
Why the safest way to modernise a legacy system is never to replace it all at once, and what one bank's £330m mistake proves about the alternative.
Every legacy system eventually tempts someone into the same plan: switch it all off, switch the new one on, and hope. In April 2018, TSB Bank tried exactly that with 5.2 million customer accounts. This playbook sets out the alternative: the strangler fig pattern, named by Martin Fowler in 2004, for modernising a legacy system component by component, safely and reversibly.
What's inside
- What actually went wrong with TSB's 2018 migration, and the FCA and PRA's own findings on why
- The strangler fig pattern, and the four phases it runs through per component
- Why reversibility, not speed, is the property that actually matters
- A side-by-side worked comparison of a big-bang migration and a strangler migration
- Common pitfalls, including the ones TSB's own regulatory findings specifically called out
- A blank canvas to map your own legacy migration
Have a question about any of these?
Have a question about any of these, or want to talk about how they'd apply to your team? Get in touch.