The library
One standard, every entry
Provenance checked, components verified against the source, a worked example, the pitfalls, and the critiques the textbooks skip. Filter by discipline, purpose, decision stage, level and format, or search by name, originator or what the tool is for.
Not sure where to start? The playbooks put the frameworks in order for the decisions that come round every year.
92 frameworks · the shelf grows one file at a time
Compare frameworks side by sideThe patterns
Not one of these? Tell me what is in front of you.
Strategy & competition
Where to compete and how to win: the industry, your position, your options.
For checking whether your business has any of the seven sources of lasting advantage, and which it could reach at its stage. Leave it alone when you need an industry-level view; this assesses one company. Instead Porter's Five Forces, for the structure everyone in the industry faces; 7 Powers assesses one company.
For sorting your growth options by whether they take existing or new products into existing or new markets, each step riskier than the last. Leave it alone when the problem is defending what you have; this generates expansion, not defence. Instead Porter's Five Forces, which reads the pressure on the position you are defending.
For translating a strategy into linked objectives and measures across finance, customers, internal processes and learning, so you steer by the drivers and not only the results. Leave it alone when there is no strategy to translate; it becomes a wall of KPIs. Instead Playing to Win, to make the choices first; the scorecard only carries choices already made.
For deciding where a portfolio's cash should go, by plotting each business on market growth and market share. Leave it alone when you are one business in one market; there is no portfolio to balance. Instead Porter's Five Forces on the one business you have; a portfolio tool with one entry has nothing to say.
For finding uncontested space by deciding which factors of competition to eliminate, reduce, raise and create. Leave it alone when you hold a defensible position in an attractive industry; exploiting it may beat abandoning it. Instead Porter's Generic Strategies, to make the most of a position worth defending rather than leaving it for open water.
For testing what your organisation is genuinely good at: the capabilities that open many markets, matter to customers and resist copying. Leave it alone when you are a young single-product company; you are still acquiring competences. Instead Lean Canvas, with cheap experiments behind it; you are still finding out what you are good at.
For deciding which business units to invest in, hold or exit, by rating each on how attractive its industry is and how strong it is in it. Leave it alone when you manage one business; there is nothing to compare. Instead Porter's Five Forces on the one business; a nine-box with one unit in it compares nothing.
For scanning the forces outside your control, political, economic, social, technological, environmental and legal, that shape the ground you operate on. Leave it alone when the question is about competition inside your industry; that is Five Forces. Instead Porter's Five Forces, for the game inside the industry; PESTEL only reads the weather.
For making a strategy an integrated set of choices: what winning means, where to play, how to win, and what capabilities and systems you need. Leave it alone when the market is too new or turbulent to choose honestly; experiment first. Instead Build-Measure-Learn; in a market too new to choose honestly, run cheap experiments before the cascade.
For judging whether an industry can pay you, from the five forces that squeeze its profits, and where you could defend a position. Leave it alone when the question is about one competitor; this analyses the structure, not the rival. Instead 7 Powers, applied to the rival, to see what power they actually hold; Five Forces reads the structure, not the player.
For choosing, deliberately, whether you compete on cost or on difference, broadly or narrowly, and not getting stuck in the middle. Leave it alone when you have not yet analysed the industry; you would be choosing a position in a market you do not understand. Instead Porter's Five Forces first; you cannot choose a position in a market you have not read.
For testing whether you actually have a strategy: a diagnosis, a guiding policy and coherent actions, or just goals and a vision statement. Leave it alone when the situation is stable and the job is execution; a fresh diagnosis manufactures drama. Instead Hoshin Kanri or OKRs, to execute the strategy you already have; a fresh diagnosis of a stable situation manufactures drama.
For building a few different, plausible futures and rehearsing what you would do in each, so the future you quietly assume stops being the only one. Leave it alone when the uncertainty is a risk you can put numbers on; forecasting and sensitivity analysis are sharper. Instead Decision Trees, when the odds can be put on the branches; scenarios are for the uncertainty that cannot.
For drawing a strategy on one page as cause and effect, from the people and processes up to the customer and the financial results. Leave it alone when the choices have not been made; a map without them is a wish list with arrows. Instead Playing to Win, to make the choices; a map drawn before them is a wish list with arrows.
For getting a team's view of strengths, weaknesses, opportunities and threats onto one page, with evidence, before deciding anything. Leave it alone when you need decisions; it organises evidence and contains no way to choose. Instead TOWS, which turns the four lists into choices; SWOT only organises the evidence.
For managing today's core, tomorrow's growing businesses and the day after's options at the same time, so the present does not starve the future. Leave it alone when disruption arrives at software speed; scheduling the future for later is how incumbents get blindsided. Instead Build-Measure-Learn now; where disruption arrives at software speed, Horizon 3 cannot wait its turn.
For turning a SWOT into strategy by pairing each strength and weakness with each opportunity and threat. Leave it alone when the SWOT underneath is thin or flattering; vague inputs give vague strategies. Instead Go back to the SWOT and make it honest; TOWS can only pair what it is given.
For finding where in your own activities cost is incurred and difference is created, because advantage lives in activities, not in the firm as a whole. Leave it alone when the question is about the industry around you; that is Five Forces. Instead Porter's Five Forces, for the industry around you; the value chain analyses your own activities.
For four tests for whether something you own is a real advantage: valuable, rare, hard to copy, and organised to exploit. Leave it alone when the question is about the outside world; this looks inward only. Instead Porter's Five Forces or PESTEL, for the outside world; VRIO only looks inward.
For mapping what serves a user need and how evolved each piece is, from brand new to commodity, so strategy becomes a shared map rather than competing opinions. Leave it alone when you need an answer this afternoon; a first useful map takes real effort and a steep learning curve. Instead SWOT or an Impact-Effort Matrix this afternoon; a Wardley map is a week's work, and worth it later.
Org, people & execution
Teams, culture, change, and getting things done through people.
For understanding why a change has not stuck with a particular person: awareness, desire, knowledge, ability, reinforcement, in that order, and which one is missing. Leave it alone when the block is structural or political; incentives and turf do not yield to individual treatment. Instead Stakeholder Mapping, because the block is in the incentives and the turf, not in the individuals.
For checking whether a team's mix of people covers the kinds of work in front of it, in nine recognisable roles. Leave it alone when you are selecting or promoting; it is a development aid and its testing does not support high-stakes judgements. Instead A structured interview and a validated assessment; this is a development aid, and it says so itself.
For managing the human side of a change already decided: the ending, the uncomfortable middle, and the new beginning. Leave it alone when you are still deciding what to change; that sits upstream. Instead Rumelt's Kernel, to diagnose and decide what to change before managing anyone through it.
For profiling an organisation's culture on two tensions, inward or outward, flexible or controlled, so it can be measured and compared. Leave it alone when you want to know why the culture is as it is; Schein digs for that. Instead Schein's Three Levels of Culture, which digs for why the culture is the way it is.
For running a coaching conversation that leaves the other person owning the answer: goal, reality, options, will. Leave it alone when the person lacks the skill; questions cannot conjure knowledge, so teach first. Instead Situational Leadership; a skill gap wants telling and showing before it wants asking.
For turning a few breakthrough objectives into aligned targets at every level, negotiated up and down rather than handed down. Leave it alone when the direction is undecided; this deploys a strategy, it does not produce one. Instead Playing to Win, to decide the direction; hoshin only deploys one you already have.
For leading a major change in the order the research says works, from urgency and a coalition to anchoring it in the culture. Leave it alone when the change is small and contained; a plan and some communication will do. Instead PDCA, if you want a shape for it; a plan and some communication will do for a change this size.
For the simplest picture of change there is: loosen the current way, move, then settle the new way so it does not snap back. Leave it alone when your organisation must change continuously; settling is the wrong ambition. Instead Kaizen and PDCA, built for change that never settles; refreezing is the wrong ambition.
For checking whether seven parts of an organisation, strategy, structure, systems, style, staff, skills and shared values, pull in the same direction. Leave it alone when the question is about markets or competitors; this looks only inward. Instead Porter's Five Forces or PESTEL for the world outside; 7S only looks inward.
For setting a few ambitious objectives a quarter, each with measurable results, published so everyone can see them and scored honestly at the end. Leave it alone when you cannot yet measure what matters; fix measurement first or the results are theatre. Instead Fix the measurement first; key results on numbers nobody trusts are theatre, however well the objectives read.
For settling who does the work, who is accountable, who is consulted and who is told, for every task that matters, with exactly one accountable name each. Leave it alone when the team is small and stable and everyone already knows; it adds admin, not information. Instead Nothing. A small, stable team that already knows who does what needs no chart.
For understanding a culture by working down from what you can see, through what people say they believe, to the assumptions that actually drive behaviour. Leave it alone when you need to compare many organisations in numbers; use a survey instrument. Instead The Competing Values Framework, a survey instrument built for comparing many organisations in numbers.
For adjusting how you lead someone on a particular task, from directing to delegating, by their competence and commitment on that task. Leave it alone when you are designing change across an organisation; this is one to one. Instead Kotter's 8 Steps or ADKAR for change across an organisation; this is one person at a time.
For plotting everyone with a stake in your work by their power and their interest, and deciding how to engage each. Leave it alone when you would do it once and file it; power and interest move, and a stale map feels like diligence. Instead The same map, redrawn at each stage; power and interest move, and a stale map only looks like diligence.
For measuring and building a team where people will flag mistakes, risks and half-formed ideas without fear. Leave it alone when you want it as a substitute for standards; safety is not low expectations. Instead Clear standards and honest accountability alongside it; safety is what lets people say the work is not good enough.
For designing software teams around the flow of work rather than history, in four team types and three ways of interacting. Leave it alone when you have no software or digital delivery; its language translates poorly outside that world. Instead McKinsey 7S, for structure in a business with no software delivery; the four team types do not translate.
For diagnosing why a leadership team underperforms, from a lack of trust at the base to a lack of attention to results at the top, and where to start the repair. Leave it alone when the problem is structural: the wrong people, an unclear mandate, no real interdependence. Instead RACI, to settle the mandate and the roles first; no amount of trust-building repairs a team that was never designed.
For a language for why a capable team is not yet delivering: forming, storming, norming, performing, and what each stage needs from its leader. Leave it alone when you want a timetable; the stages describe tendencies, not a schedule. Instead A project plan for the timetable; Tuckman describes tendencies, and "storming week" is not a date.
Decision-making & prioritisation
Choosing between options, structuring a problem, and preparing to decide.
For preparing for a negotiation by knowing your best alternative to a deal, the worst deal you would still accept, and where the two sides overlap. Leave it alone when the other side holds all the cards and knows it; improve your alternative first. Instead Improve your alternative first, or decide whether to negotiate at all; no technique makes up for a bad walk-away.
For working out what kind of problem you are facing, so you decide the way the problem needs rather than the way you are comfortable with. Leave it alone when you want a label to file and forget; situations move. Instead Nothing to file; come back to it when the situation moves and you need to know what kind it has become.
For choosing between options by agreeing what matters, weighting it, and scoring each option, so the reasoning can be seen and argued with. Leave it alone when one thing dominates; if price is everything, a blend only hides that. Instead Say which one thing matters and decide on it; Cost-Benefit Analysis if it is money, a Risk Matrix if it is safety.
For laying out a decision with branches for what you control and what you do not, with rough odds and payoffs, so you can see which first move is worth most. Leave it alone when nobody can put even a rough number on the odds; it will manufacture precision. Instead Scenario Planning, which works without odds you cannot defend.
For sorting your work by urgent and important into do, schedule, delegate or drop. Leave it alone when the tasks depend on each other; a project needs a plan, not a grid. Instead A plan with the dependencies in it; tasks that wait on each other need sequencing, not sorting.
For finding the quick wins by sorting ideas by how much they would do against how much they would cost. Leave it alone when the decision is big or hard to undo; two gut-feel axes are too crude for that. Instead Cost-Benefit Analysis, with a Risk Matrix beside it; a big or irreversible call deserves more than two gut-feel axes.
For breaking a big question into parts that do not overlap and leave nothing out, so you can size them and work on the few that matter. Leave it alone when the answer is already obvious; building a tree to reach it is theatre. Instead Act on the obvious answer; a tree built to reach what everyone already knows is theatre.
For agreeing before a deadline what must be delivered, what should, what could, and what will not this time. Leave it alone when you need a rank order; this sorts into four bins and says nothing about the sequence inside them. Instead RICE, which ranks; MoSCoW only sorts into bins.
For deciding faster than the situation changes: observe, orient, decide, act, and round again. Leave it alone when it is a one-off, high-stakes call with time to think; use a decision tree or a cost-benefit analysis. Instead Decision Trees or Cost-Benefit Analysis; a one-off, high-stakes call with time to think wants analysis, not tempo.
For before you commit, imagining the plan has already failed and having everyone write down why, in silence; it costs an hour. Leave it alone when the decision is already signed; it produces regret, not improvement. Instead A Risk Matrix on the signed decision, to manage what you can no longer avoid.
For writing a recommendation the reader can act on: the answer first, then the reasons, grouped so they hold up. Leave it alone when the thinking is not finished; answer-first dresses up a conclusion you do not have. Instead Issue Trees, to finish the thinking; answer-first comes after there is an answer.
For ranking competing ideas by reach, impact, confidence and effort, so quiet, valuable work can beat the loudest request. Leave it alone when the question is strategic rather than a list; it ranks, it does not set direction. Instead OKRs or Three Horizons, to set the direction; RICE only ranks a list against one.
For running a discussion where everyone looks at the facts, the risks, the upside and the new ideas together, one mode at a time, instead of taking sides. Leave it alone when you need numbers compared against criteria; use a decision matrix. Instead A Decision Matrix, for the arithmetic; the hats structure the talk, not the evaluation.
Operations & process
Fixing a process that keeps going wrong, and finding the bottleneck.
For fixing a process that keeps producing defects, in five disciplined steps from defining the problem to holding the gain. Leave it alone when you already know the cause and the fix; replace the valve. Instead Replace the valve. A known cause with a known fix needs a work order, not a five-phase project.
For getting every plausible cause of a problem out of people's heads and onto one organised diagram, ready to test. Leave it alone when the cause is known; drawing a skeleton round a solved problem is decoration. Instead PDCA, to fix the cause you already know and check the fix held.
For getting from a symptom to the cause underneath it by asking why of each verified answer, until you reach a fix that will hold. Leave it alone when the failure has several interacting causes; it forces a single path. Instead A Fishbone Diagram, which holds several interacting causes at once.
For improving the way things are done in small, cheap, constant steps that everyone takes part in, locking each gain in as the new standard. Leave it alone when the process or the business model is fundamentally wrong; refining the wrong thing entrenches it. Instead The Business Model Canvas, to redesign it; refining the wrong thing only entrenches it.
For finding the few causes behind most of the trouble, by ranking the losses and seeing where the weight sits. Leave it alone when the causes are all about the same size; forcing a vital few onto a flat spread invents a priority. Instead A Fishbone Diagram; when the causes are all the same size, look at how they connect rather than which is biggest.
For testing a change on a small scale, checking the result against what you predicted, and keeping what works before going round again. Leave it alone when you are in an emergency; contain it first, improve it later. Instead Contain it first. An outage or a safety incident needs stopping, not studying; come back to PDCA when it is stable.
For agreeing the boundaries of a process on one page, suppliers, inputs, steps, outputs and customers, before anyone argues about the detail. Leave it alone when you already need the detail, the times, the queues, the rework; use value-stream mapping. Instead Value Stream Mapping, which has the times, the queues and the rework that SIPOC leaves out.
For finding the one bottleneck that sets how much the whole system can do, and pointing everything at it. Leave it alone when there is no stable bottleneck, because demand is low everywhere or the constraint moves week to week. Instead Value Stream Mapping, to level the flow when there is no stable bottleneck to elevate.
For drawing every step, queue and hand-off a product passes through, so you can see how little of the time adds value and design a leaner flow. Leave it alone when the problem sits inside one step; study that step directly. Instead A Kaizen event on the one step; a door-to-door map is for problems that sit between steps.
Innovation & product
New ventures and new products, from the idea to the first customers.
For turning an idea into the smallest thing a customer can react to, measuring what they do, and learning whether to change course or carry on. Leave it alone when the requirement is already known; iterating towards a fixed specification is waste dressed as learning. Instead A plan and a delivery team; when the requirement is known, build it and test it against the specification.
For seeing a whole business on one page, from who it serves to what it costs, so it can be questioned and redrawn. Leave it alone when the question is about competitors or the industry; the canvas has no box for rivals. Instead Porter's Five Forces, for the rivals the canvas has no box for; PESTEL for the weather around them.
For solving a problem by starting with the people who have it, defining it properly, and prototyping before you commit. Leave it alone when everyone agrees what to build; prototyping the question again is theatre. Instead PDCA, to build what everyone has agreed and check it as you go.
For understanding why well-run companies lose to cheaper, simpler rivals, and seeing it coming. Leave it alone when the new rival came in at the top with a better, dearer product; that is a different fight. Instead 7 Powers, to see what defence you hold against a better, dearer rival coming in at the top.
For finding the right problem before finding the right solution, in two rounds of widening and narrowing. Leave it alone when the problem is already defined and tested; start on the solution half. Instead Build-Measure-Learn, straight to the solution half, since the problem is already settled.
For a one-page plan for a new venture that puts the riskiest assumptions, the problem, the solution, the unfair advantage, where you can see and test them. Leave it alone when the business is established with real partners and operations; the full canvas fits better. Instead The Business Model Canvas, which keeps the partners, channels and operations an established business has.
For matching what one group of customers is trying to do, and what frustrates them, against what you offer and how it helps. Leave it alone when you are designing the whole business; this covers two of its nine blocks. Instead The Business Model Canvas, which covers the other seven blocks a whole business needs.
Marketing & customer
Who to sell to, what to say, and what to charge.
For auditing an advert, a page or a pitch against the oldest checklist in marketing: does it get attention, hold interest, create desire, and ask for action? Leave it alone when you want to know how customers actually buy; real buying loops and stalls. Instead Customer Journey Mapping, which follows how people actually buy, loops and stalls included.
For getting a genuinely new product from the enthusiasts who will try anything to the mainstream buyers who want references, by winning one narrow segment first. Leave it alone when the product asks no change of behaviour from the buyer; there is no chasm to cross. Instead STP; if the product asks no change of the buyer, it is ordinary targeting, not a chasm.
For walking the customer's whole experience of you, step by step, with what they do, think and feel, to find where it breaks. Leave it alone when you have no customer research and no time to get any; a map built from opinion is guesswork with polish. Instead Jobs to be Done: a handful of real customer conversations is the research the map needs.
For checking that your product, price, place and promotion add up to one coherent offer for a chosen customer. Leave it alone when you have not chosen the customer yet; run the targeting first or you will document a programme aimed at nobody. Instead STP first; choose the customer, then set the mix for them.
For dividing a market into groups, choosing which to serve, and giving them a clear reason to prefer you. Leave it alone when the market is too small or too uniform to divide; one well-made offer may serve it better. Instead The Value Proposition Canvas, to make one well-made offer for a market too small or too uniform to divide.
For setting a price from the value you create for a chosen customer rather than from your costs, and testing what they will pay before you launch. Leave it alone when you sell a commodity with a public price; there is no value difference to price. Instead Break-even Analysis and Unit Economics; on a commodity with a public price, the job is cost, not value.
Finance
Can we afford it, and is the business healthy?
For finding how much you have to sell before you stop losing money, and how much room the plan has to be wrong. Leave it alone when costs cannot be split into fixed and variable, as in a business where every job is priced from scratch. Instead Unit Economics, job by job: what each one earns after what it costs to deliver.
For seeing when you will run short of cash, month by month, even while the profit figures look fine. Leave it alone when you want to know whether you are profitable; this measures cash, not profit. Instead Financial Ratio Analysis, which reads profit from the accounts; the forecast only tells you whether you can pay the bills.
For putting every cost and benefit of a proposal into today's money and comparing the totals. Leave it alone when what matters most cannot be priced, such as reputation or safety; forcing it produces confident nonsense. Instead A Decision Matrix, where reputation and safety can stay in their own units instead of being forced into pounds.
For reading a set of accounts to answer two questions: can we afford this, and is the business actually healthy? Leave it alone when the business has traded for less than a year and there is nothing clean to compare. Instead Cash Flow Forecasting; in the first year the question is whether you will still be trading, not how the ratios compare.
For putting a rough, defensible number on a private business with an earnings multiple and a simple discounted cash flow, to open a serious conversation. Leave it alone when a transaction, a tax event or a dispute is on the table; that needs a qualified valuer. Instead A qualified valuer or accountant; a transaction, a tax event or a dispute needs a figure that carries weight.
For checking whether each sale and each customer actually makes money, before you scale and multiply the loss. Leave it alone when you have too little history to know how long customers stay; three months of data is a guess with a decimal point. Instead Break-even Analysis and Cash Flow Forecasting until you have a year of customers; lifetime value needs a lifetime.
Customer & product
What customers are actually trying to do, and whether you help.
For working out what a customer is really trying to get done when they buy from you, and what pushes them to switch or holds them back. Leave it alone when you need numbers across a whole market; this is a handful of good conversations, not a survey. Instead The Kano Model, surveyed properly, when you need numbers across a whole market.
For sorting what your product offers into the things customers expect, the things they will pay for more of, and the things that delight them. Leave it alone when you cannot ask real customers; guessed answers are the very thing it exists to replace. Instead Jobs to be Done, a few real conversations, when you cannot survey; a Kano drawn from guesses measures nothing.
For one question that gives you a single loyalty number to track over time: how likely are you to recommend us? Leave it alone when you want it to be the only measure of customer health; it is one indicator, not the verdict. Instead Put it beside Customer Journey Mapping and the Kano Model; one number is an indicator, not a verdict.
For keeping product discovery honest by linking the outcome you want to the customer opportunities behind it, the ideas for each, and the tests that decide. Leave it alone when there is nothing to discover, such as a legal requirement or a plain fix. Instead Just do it; a legal requirement or a plain fix needs delivery, not discovery.
For drawing a service end to end, what the customer does and everything behind the scenes that makes it happen, to find where it breaks. Leave it alone when you only need the customer's view; a journey map is lighter. Instead Customer Journey Mapping, which is lighter and gives the customer's view on its own.
Governance
Who decides, who checks, and who answers for it.
For giving a board a structure for overseeing AI, govern, map, measure, manage, inside the governance it already has. Leave it alone when you have a legal AI obligation; this is a leadership framework, not compliance advice. Instead Legal advice on the obligation itself; this is a leadership framework, not compliance.
For working out, in a family firm, who is speaking as family, who as owner and who as manager, so a row about who is right becomes a question about which line is being crossed. Leave it alone when there is no family in the ownership; nothing here is for you. Instead RAPID Decision Roles or RACI; without a family in the ownership, the question is who decides, not which circle they sit in.
For deciding which of your impacts on people, the environment and the economy matter enough to report on and act on. Leave it alone when the law already tells you what to report; this is the voluntary version. Instead Your regulator's own process, with advice; this is the voluntary version.
For naming, for one big decision, who recommends, who must agree, who gives input, who decides and who carries it out, so one person owns the call. Leave it alone when decisions get made fine and the problem is their quality; this fixes who, not how well. Instead A Decision Matrix or a Pre-mortem; RAPID fixes who decides, not how well.
For separating the people who run the risks from the people who check them, and both from the people who assure the board. Leave it alone when the organisation is small enough that all three lines would be the same two people; keep the principle, drop the structure. Instead RACI, to keep doing and checking in separate hands when the three lines would be the same two people.
Risk
What could go wrong, how badly, and what to do first.
For asking of every part of a design or process how it could fail, how badly, how likely, and whether you would catch it, then fixing the worst first. Leave it alone when the risk is strategic or systemic; this works part by part. Instead A Risk Matrix for the systemic risks, Scenario Planning for the strategic ones; FMEA works part by part.
For forecasting a project from what actually happened to projects like it, rather than from the plan in front of you, to correct the optimism. Leave it alone when nothing comparable has ever been done; a forced comparison gives confident numbers with no foundation. Instead Scenario Planning and staged commitment, when nothing comparable has ever been done.
For putting every risk on a grid of how likely and how bad, so attention and money go to the ones that are both. Leave it alone when the decision turns on precise quantities, such as reserving or insurance; a grid cannot carry that weight. Instead Quantitative risk analysis by someone qualified; reserving, insurance and safety cases need numbers, not a grid.