What a framework is for
A framework is not a method and not a theory. It is a stable way of dividing a problem into parts so that the parts can be examined one at a time and nothing obvious is forgotten. The marketing FunnelModel of the customer journey stages. divides a buying process into stages; the marketing mix divides an offer into levers; a planning model divides a plan into sections. In every case the value is the same: a shared vocabulary, a checklist of things to consider, and a structure that survives the person who first drew it on a whiteboard.
Frameworks also hide things. A funnel hides the people who move backwards or sideways; a stage model hides the time spent in each stage; a measurement model hides the data it cannot collect. Each page in this library therefore says what the framework makes visible and what it leaves out, because the second list is where most mistakes with frameworks are made. A framework is a lens to look through, not evidence of how the world behaves, and the moment a team starts defending the model against what the data shows, the model has stopped being useful.
The frameworks here are described as practitioners use them. A framework is attributed only where a source the page cites does so; otherwise the page names the framework, describes it without attribution and says nothing about its history. No page quotes adoption figures or uplift claims, because a framework's worth cannot be measured that way.
The seven groups
The library sorts frameworks by the kind of decision they serve. A framework can be useful in several places; the group is where it is used most.
| Group | What the frameworks in it decide | The question they answer |
|---|---|---|
| Strategy | Where to compete, for whom, and with what offer | "Which market, which segment, which position, which levers?" |
| Messaging | What a message must do and in what order | "What does this ad, page or email need to earn before it asks for action?" |
| Funnel & Journey | How people move towards and beyond a purchase | "What does someone need at each stage, and what do we provide there?" |
| Planning | How a strategy becomes a programme of work | "What goes in the plan, in which order, and who owns each part?" |
| Measurement | How to know whether the work is succeeding | "Which metric, which KPI, which model of credit?" |
| Growth | How acquisition can compound rather than restart | "What mechanism brings the next customer without a new budget?" |
| Experimentation | How to test a change honestly | "What do we test, in what order, with what decision rule?" |
The groups also form a rough sequence. Strategy frameworks come first because everything downstream depends on who the customer is and why they would choose this offer; messaging and journey frameworks turn that choice into things people see; planning frameworks organise the work; measurement frameworks say whether it worked; growth and experimentation frameworks improve it. The pages beneath this introduction are listed automatically under these seven labels, so a new framework appears here as soon as it is published.
Choosing a framework for a decision
Start from the decision in front of you, not from the framework you know best. The table routes the most common decisions to a starting framework and the one to reach for next.
| Decision in front of you | Start with | Then |
|---|---|---|
| Choosing which customers to serve and how to be different | STP, with Jobs-to-be-Done for the demand behind the segments | The positioning statement to write the choice down |
| Choosing a direction for growth: new markets, new products or both | The Ansoff matrix to compare the moves by how much is new | Five forces for the pressure the chosen market puts on margin and acquisition cost; STP for whom to serve in it |
| Checking an offer is complete before launch | The marketing mix | SWOT and PESTEL for the situation the offer enters |
| Pricing a range of versions for different buyers | Good-better-best tiering for the job of each tier and the fences between them | The marketing mix to check each tier against the rest of the offer |
| Building a brand that buyers think of and can find when they buy | Mental and physical availability | Distinctive brand assets for the cues that identify the brand without its name; double jeopardy for why penetration comes before loyalty |
| Writing or reviewing an ad, landing page or email | AIDA | The Value Proposition Canvas if the message keeps failing at desire |
| Turning product features into reasons to buy | Features, advantages, benefits (FAB) and its so-what test | AIDA for the order the page, ad or email runs in |
| Testing whether a PR idea or announcement is news | News values | The inverted pyramid to put the conclusion first in the release, email or report that carries it |
| Planning content and channels by stage | TOFU, MOFU and BOFU or See-Think-Do-Care | Customer journey mapping when stage labels are too coarse |
| Deciding how to treat existing customers by how they buy | RFM segmentation to score them on recency, frequency and monetary value | STP for retention to decide which segments to serve at all |
| Writing an annual or quarterly marketing plan | SOSTAC for the plan's sections | RACE for objectives and KPIs by stage; media planning for the budget split |
| Keeping part of a budget for promising and experimental work | The proven, promising and experimental budget split and its rules for moving money between tiers | Media planning for the split across channels |
| Prioritising mixed work or choosing between a few options | The impact-effort matrix to sort audit findings, fixes and requests | The weighted scoring matrix when choosing between known options such as markets, tools or agencies |
| Planning under uncertainty or staging a launch | Scenario planning for a few plausible futures with triggers | Phase-gate reviews to commit in stages against criteria agreed in advance |
| Sharing ownership of a launch, an approval or a campaign | The RACI matrix with one accountable owner per decision | The approvals and compliance reviews guide for routing sign-off by risk |
| Deciding what to measure | The measurement plan | The KPI tree, AARRR or North Star metric and OKRs for the metric hierarchy |
| Deciding how to credit channels | Attribution models | The attribution guides for the implementation |
| Writing report commentary that leads to action | What, so what, now what | Control charts to tell real change from noise before reporting a movement |
| Finding the cause of a failure, an outage or a run of complaints | The five whys until the cause is one that can be changed | The campaign post-mortem template to record the cause and the change agreed for next time |
| Turning a metric into a target or an incentive | Goodhart's law for the guardrails that keep the measure informative | The KPI tree to set the target alongside the measures around it |
| Finding how acquisition can compound | Growth loops | The flywheel for the organisational version |
| Finding the one step that limits growth or output | The theory of constraints | AARRR to locate the funnel stage where the constraint sits |
| Deciding which buyers to reach first with something new | Diffusion of innovations for what each adopter group needs | STP for launches to pick the first segment |
| Deciding how much evidence a decision needs | One-way and two-way doors to sort choices by how reversible they are | The experimentation loop for the decisions worth testing |
| Prioritising and running tests | The experimentation loop and ICE | The conversion optimisation cluster for the test programme itself |
| Scheduling work back from a fixed launch or event date | Backward planning for a latest start on every step the date depends on | Phase-gate reviews to commit in stages; the RACI matrix for who owns each step |
| Deciding where to concentrate effort or budget | The Pareto principle to rank customers, products, partners or campaigns by the value they carry | The impact-effort matrix to sort the work; RFM segmentation to score the customers |
| Setting objectives and spend as a product ages | The product life cycle, with the stage read from the business's own sales, search and stock data | Diffusion of innovations for what each adopter group needs; the Ansoff matrix to compare the next moves by how much is new |
| Comparing options with uncertain payoffs | Expected value to weight each outcome by its probability | Scenario planning for a few plausible futures with triggers |
| Judging how strong a claim of effect is | The hierarchy of evidence to rank the proof from anecdote to randomised test | The experimentation loop for the claims worth testing |
| Weighing what buyers say against what they do | Stated and revealed preference | Jobs-to-be-Done for the demand behind the choices |
| Removing an old field, report, rule or tool | Chesterton's fence to find out why it exists and who still relies on it | One-way and two-way doors to sort the removal by how reversible it is |
| Writing claims that win the click without losing the customer | Expectation-disconfirmation to judge each promise against the experience | Customer journey mapping to follow the promise through the journey |
| Reading research or reports that hear only from those who stayed | Survivorship bias to name the missing group and bring it back | Goodhart's law for the guardrails that keep a measure informative |
| Choosing between uses of the same money or time | Opportunity cost to judge each option against the best alternative, not against doing nothing | The weighted scoring matrix to compare the alternatives against agreed criteria |
| Governing the team's use of AI | The NIST AI Risk Management Framework and its four functions: govern, map, measure and manage | The RACI matrix for who is accountable for each use |
| Deciding which deviations from a plan need attention, and who acts on them | Management by exception to agree a tolerance for each line in advance and escalate only what breaks it | Control charts to set a metric's normal range from its own history; the RACI matrix for who acts on each exception |
Two frameworks from different groups are usually better than two from the same group. A plan written in SOSTAC with RACE inside it covers more than a plan written in two planning models, because the second model adds a different kind of question rather than a second version of the same one.
A meal-kit subscription company choosing where to start
A meal-kit subscription company with a marketing team of three and an annual budget of £240,000 has never written a plan. The team's first instinct is a measurement framework, because the board asks about numbers. Working through the table, they realise the open decision is upstream: they have not decided which of their two segments, busy families and single professionals, the year is for. They run STP first, choose families on the strength of a 14-month average subscription length against 5 months for professionals, and write a positioning statement. Only then do they draft the plan in SOSTAC, set RACE objectives of 9,000 new households and a 70 percent three-month retention rate, and build a KPI tree beneath those two figures. The measurement framework came last, and it was the right order.
How each framework page is organised
Every framework page follows the same shape so that a reader who knows one can read any other quickly. It opens with a short summary of what the framework is and what it is for. Its origin is stated only where a source the page cites documents it, and only in that source's terms; otherwise the framework is described without attribution and the page spends no words on history. The framework's components appear as a table, because most frameworks are lists with a structure, and a table shows the structure. A section on applying it explains how a practitioner uses it in a real planning or review session, and which inputs it needs.
A worked example, always labelled illustrative, shows the framework applied to a concrete scenario. The example is the only place on a page where figures appear, so that no reader mistakes an invented scenario for a benchmark. A section on limits and criticisms says what the framework hides and where it is commonly misused. Related frameworks are linked with a sentence on how they combine. Each page ends with common mistakes, three short answers to the questions people most often ask about the framework, and the group label that places it in the table above.
Frameworks used across this site
The frameworks are not a separate library from the rest of the site; they are the thinking behind it. Guides declare the frameworks they use in their metadata, so each framework page shows where it is applied, and the applied guides point back here for the reasoning. The digital marketing strategy hub uses the strategy and planning groups directly: segmentation, positioning, SWOT and the planning models all appear there as working steps rather than as frameworks to be studied. The marketing funnel basics guide is the gentle introduction to the funnel and journey group, and the right place to begin if the stage vocabulary is new.
The measurement group underpins the analytics measurement guides: a KPIA metric tied directly to an objective. is only a KPI because a framework tied it to an objective, and every AttributionAssigning conversion credit to touchpoints. report is an attribution model applied, whether or not the model was chosen deliberately. The experimentation group is the method behind the conversion rate optimisation cluster, where the loop of hypothesis, test and decision runs continuously. For how the author uses these models in client and in-house work, the marketing frameworks resource page describes the working practice, while the pages in this library explain the frameworks themselves.
