---
name: needs-motivation-and-jobs-to-be-done-analysis
description: >
  Establishes what an audience is actually trying to achieve, separating wants from
  needs, motivations from barriers and stated barriers from operative ones, by
  eliciting needs indirectly rather than asking people what they need, laddering
  from attribute to consequence to value, and treating jobs-to-be-done as a
  disciplined framing rather than a validated theory. Use for "what do customers
  really need", "unmet needs analysis", "jobs to be done", "why do people choose
  this", "customer motivations", "what is stopping them", "needs-based
  segmentation input", "laddering", "the struggling moment", "functional and
  emotional needs".
category: 09 Segmentation and Audience Understanding
ref: 09.05
tier: 1
inherits: [K2, K3, K4, K5]
---

# Needs, Motivation and Jobs-to-be-Done Analysis

## 1. One-line description
Establishes what people are trying to achieve and what is stopping them, by distinguishing five things that are routinely conflated (want, need, motivation, barrier and job to be done), eliciting needs through behaviour, struggle and consequence rather than by asking people to name them, and reporting each need claim with the evidence behind it and the confidence it has earned.

## 2. What this skill is used for

**The research problem it solves.** Almost every commercial brief contains the phrase "customer needs", and almost none of them defines it. In practice the word is used for five different things: a want (something a person would like), a need (something whose absence causes a real problem), a motivation (the goal that makes a behaviour make sense), a barrier (what prevents the behaviour) and a job to be done (the progress the person is trying to make, of which the product is one possible solution). These are not synonyms, they lead to different decisions, and confusing them is expensive. The second problem is elicitation. Asked what they need, people answer with a solution, usually a slightly better version of what they already have, because articulating a need requires a level of introspective access nobody has on demand. So a needs study conducted by asking about needs produces a feature list. The third problem is that "unmet need" has become a phrase that markets an idea rather than describing evidence, and a preference somebody voiced once is routinely elevated to an unmet need because the phrase makes the finding sound bigger. This skill supplies the definitions, the indirect elicitation methods, the laddering discipline and its interpretive limits, an honest treatment of the jobs framing, and the evidence standard that separates a need from a wish.

**Where it sits.** Analysis and synthesis, primarily on qualitative material, frequently in combination with behavioural evidence. It supplies the needs dimensions that segmentation is built on and the goal, job, barrier and unmet-need elements that personas carry.

**Typical use cases.**
- Establishing what an audience is trying to achieve, ahead of a product, service or proposition decision.
- Identifying and evidencing unmet needs, and distinguishing them from preferences and wishes.
- Producing needs-based basis variables for a segmentation.
- Understanding why a service is not adopted, by separating stated barriers from operative ones.
- Analysing switching, using the struggling moment and the forces acting on the decision.
- Auditing a needs framework or a jobs statement that somebody else produced, particularly one that reads as a feature list.

**Who uses it.** Qualitative researchers and mixed-method leads running needs and motivation work; product, service and innovation teams who need a stable frame that outlives a particular solution; strategists building propositions; UX researchers working on adoption and abandonment; and anyone reviewing AI-generated needs analysis, where the output reliably contains fluent, plausible, unevidenced needs because the concept is easy to write and hard to check.

## 3. When to use it

- A product, service or proposition decision depends on what people are trying to achieve, not on what they say about the current offer.
- Adoption is lower than expected and the reasons given by non-users do not explain it.
- A needs-based segmentation is planned and the needs dimensions have to come from somewhere defensible.
- The organisation talks about customer needs and cannot say what evidence any of them rests on.
- Switching behaviour needs explaining, and the moment of switch is available to study.
- A category is being redefined and the question is what people would use instead if the category disappeared.
- An innovation pipeline is full of features and empty of problems.
- You are auditing a jobs-to-be-done framework that has drifted into a list of product improvements.

## 4. When NOT to use it

- **The only available evidence is people answering direct questions about their needs.** A "what do you need from a provider" item produces a solution list ranked by familiarity. It is not useless, and it is not a needs analysis. Report it as stated preference, name what it is, and either obtain evidence of behaviour, struggle or consequence, or stop. Converting a stated preference list into a needs framework is the commonest way this work fails, and the output is indistinguishable from real needs analysis at the point of reading.
- **The question is what people do, not what they are trying to achieve.** Behaviour, frequency, repertoire and occasion are **09.03 Behavioural Profiling**, and needs analysis without a behavioural baseline tends to float free of anything checkable.
- **The question is which need matters most commercially.** Prioritisation and sizing are **08.05 Insight Prioritisation and Sizing**. A needs analysis that arrives pre-prioritised has usually had the prioritisation imported from a stakeholder view.
- **The purpose is to justify a solution that already exists.** Where a team has built something and wants the need that explains it, this becomes reverse-engineering, prohibited by K4 §4.2. Offer instead an explicit assessment of the proposed need against all the evidence, including the evidence that cuts against it, reported either way.
- **The evidence base is too thin to distinguish a need from a preference.** Two participants mentioning something is a lead, not an unmet need. Where the corpus will not support the distinction, say so and report the material as a hypothesis with the validation named, per K3 §4.3. A short, honest needs list beats a long speculative one, and it is the difference between a document a product team can act on and one they will quietly ignore.
- **The category is one where the job framing does not apply usefully.** Jobs-to-be-done is a framing, and like all framings it fits some situations and distorts others. Habitual, low-involvement, socially embedded and pleasure-driven consumption often has no identifiable "job" that a person is hiring a product to do, and forcing one produces a strained abstraction. Say the framing does not fit, and use motivations and occasions instead.
- **Cultural reading is required and unavailable.** What a need means, what may be said about it and what a shrug signifies are culturally located, per K5 §2.2. Needs analysis from translated transcripts without a reader who knows the context produces confident and wrong interpretation, especially around social and status dimensions.
- **The topic is sensitive and the analysis would expose participants.** Needs work often reaches financial difficulty, health, capability and shame. Where the analysis would make individuals identifiable or would report a group's needs in a way that could be used against them, that is **13.05 Research Ethics and Consent Design** first.

## 5. Required inputs

**Required.** Without these the skill cannot run. If absent, ask. If the first is absent, stop.

- **Evidence about behaviour, struggle or consequence.** Depth interviews, ethnography, service records, complaint logs, abandonment data, switching accounts, diary material. Something that shows what people did and what happened, rather than only what they say they want. **If the only input is a direct needs question, do not produce a needs framework.**
- **Participant identifiers and the ability to trace every claim to source**, per K2 §4.2. A need without a traceable base is an assertion.
- **The decision the analysis informs.** Determines the level of abstraction at which needs should be stated, which is the most consequential judgement in the whole method.
- **The category and context definition.** What counts as being in this situation, because a job statement is meaningless outside the situation it describes.

**Optional, and what each one adds.**

- **Behavioural or transactional data:** allows a claimed need to be checked against what people actually did, which is the strongest available test and the one that kills most spurious needs.
- **Accounts of the moment of change** (switching, first purchase, abandonment, cancellation): allows the struggling moment and the forces framing to be used, which is where jobs analysis is at its most reliable.
- **Non-users and rejecters:** the strongest source of barrier evidence, and routinely absent from samples drawn from customer lists, which systematically excludes the people whose barriers were operative.
- **Competitive and substitute usage, including non-category substitutes:** reveals what else people hire to do the same job, which is usually the most strategically useful output.
- **Quantitative validation of the need statements:** allows prevalence and importance to be estimated, converting a qualitative framework into something sizeable.
- **Complaint, support and service-failure records:** need evidence in its most unfiltered form, because these describe consequences that actually occurred.

## 6. Questions to ask before starting

1. **What decision does this inform, and at what level of abstraction is it useful?** "Wants to feel in control of their money" and "wants to see the balance before the payment leaves" are both true and only one of them is actionable for a given decision. *Default if unanswered:* state needs at the level at which the organisation can act, and show the ladder above and below so the level can be changed.
2. **What evidence exists beyond what people said they wanted?** Determines whether a needs analysis is possible at all. *Default:* if nothing exists, say so and report stated preferences as stated preferences.
3. **Does the sample include people who did not adopt, who left, or who chose something else?** Determines whether barrier findings are credible. *Default:* state that barrier evidence from users describes the barriers users overcame, which is a different population from the one that did not.
4. **Is this a category where the job framing applies?** Determines the whole structure of the analysis. *Default:* test the framing on two or three cases before committing to it, and abandon it if the job statements are strained.
5. **Is there a moment of change available to study?** Determines whether the forces framing can be used. *Default:* work from current behaviour and struggle, and note that the switching evidence is retrospective and reconstructed.
6. **What would count as evidence that a need is unmet rather than simply unstated?** Sets the standard before the analysis rather than after. *Default:* require evidence of consequence: a workaround, an abandonment, a complaint, a substitute, a cost paid.
7. **Are any needs likely to be in conflict with each other?** Determines whether the output should be a list or a set of tensions. *Default:* look for conflicts explicitly, since a needs list with no tensions in it has usually been tidied.

## 7. Step-by-step methodology

**The position this method takes.** The words in this area are used loosely and expensively, so the method begins by fixing them and holds them throughout. **A want** is something a person says they would like. **A need** is something whose absence produces a consequence they can point at. **A motivation** is the goal that makes a behaviour rational from where the person stands. **A barrier** is what prevents or slows the behaviour. **A job to be done** is the progress a person is trying to make in a situation, for which the product is one candidate solution. The practical difference between a want and a need is evidence of consequence, and asking for that evidence removes most of what usually appears on a needs list.

**1. Establish the situations before the needs.** A need is meaningless without the situation in which it arises, so map the situations first: when does this arise, what is the person doing, what constraints are on them, what has to be true for this to matter. Work from behavioural and occasion evidence per **09.03** where it exists. Needs stated without situations become abstractions that fit everyone and guide nobody ("wants convenience"), and the fastest way to produce a useless needs framework is to skip this step. *Correct result:* a situation inventory, each entry describing a recognisable circumstance with its constraints, before any need is named.

**2. Elicit indirectly, because people report solutions rather than needs.** Six lines of evidence work, and none of them is a direct question. **Workarounds:** what has the person built, improvised or done twice because the offer does not do it? A workaround is the strongest single indicator of an unmet need, because somebody spent effort on it. **Abandonment and failure points:** where did they stop, and what happened next? **Substitutes, including non-category ones:** what else gets used for this, which reveals what the job actually is. **Consequences:** what went wrong when it was not available, and what did that cost them in money, time, effort or standing? **Compensating behaviour:** over-ordering, over-checking, keeping a spare, arriving early, asking someone else to do it. **Emotional markers in the account:** irritation, embarrassment, relief, defensiveness, which mark where something is at stake rather than what the something is. Where a direct question has been asked, treat the answer as evidence about how the person describes their situation, not as the need itself. *Correct result:* a needs evidence table with the elicitation route recorded for each candidate need, and no candidate resting solely on a direct answer.

**3. Ladder from attribute to consequence to value, and stop before the abstraction becomes universal.** Laddering moves from what the person mentions (an attribute of the product or service) to what it does for them (the functional consequence) to why that matters (the personal value). "I want the local branch" → "so I can ask someone and know it's been done properly" → "so I don't lie awake wondering whether it went through". Two disciplines make this reliable. **Ladder from a specific instance, not from a general opinion**, because a general opinion produces a general ladder that fits anyone. **Stop climbing when the rung stops being distinctive**, which is usually at the second or third step: keep going and every ladder in every category terminates in security, freedom, belonging or self-esteem, which is true, universal and useless for any decision. Laddering is interpretive, so it carries a specific risk: the analyst supplies the rungs the participant did not. Record which rungs came from the participant's own words and which are the analyst's inference, per K2 §3.1, and where the corpus allows, check whether the same ladder appears independently in other participants. *Correct result:* laddered chains with participant-sourced rungs distinguished from inferred rungs, terminating at the highest distinctive level rather than the highest possible one.

**4. Write job statements, and be honest about what the framing is.** A job statement describes the progress a person is trying to make in a situation, in a form that does not name a solution: "when I have just been told I need a procedure, help me work out what happens next, so I can tell my family something definite". The discipline it buys is real and it is the reason to use it: **the job is stable while solutions change**, so a framework built on jobs survives a technology shift that would invalidate one built on product attributes, and it forces the analyst to notice non-category substitutes. The honesty required is equally real: jobs-to-be-done is a framing, not a validated theory. It has no agreed operational definition, no established measurement, and no body of evidence establishing that people experience their behaviour this way. It organises evidence usefully in some categories and strains badly in others, particularly habitual, hedonic and socially embedded consumption. Say which it is doing here. A job statement is an interpretation of evidence and is labelled as one, with the evidence beneath it, never presented as something a participant said. *Correct result:* job statements in situation-plus-progress form, each traceable to evidence, with a stated assessment of how well the framing fits this category.

**5. Separate the functional, emotional and social dimensions of each job.** Most jobs have all three and most analyses report only the first. **Functional:** the practical task to be accomplished. **Emotional:** how the person wants to feel, or to avoid feeling, in that situation. **Social:** how they want to be seen, or to avoid being seen, by specific others. The social dimension is the most commonly missed and often the most decisive, because it explains behaviour that looks irrational when only the functional job is considered: a person who chooses a slower route because the faster one requires admitting they do not understand something. Evidence for emotional and social dimensions has to be as real as for functional ones. An inferred feeling is an interpretation, per K4 §3.4, and must be marked. *Correct result:* a three-dimension breakdown per job, with each dimension either evidenced or marked as inferred with its confidence.

**6. Analyse the moment of change with the forces framing, where a moment exists.** Where the evidence includes switching, adoption, abandonment or cancellation, four forces act on the decision and all four should be evidenced separately: the **push** of the current situation (what became intolerable), the **pull** of the new option (what attracted), the **habit** holding the person in place (inertia, sunk cost, familiarity, arrangements already made), and the **anxiety** about the new option (what might go wrong, what would be lost, what has to be learned). The value of this framing is that the last two are usually stronger than the first two and are routinely absent from analysis, so interventions get designed to increase pull when the operative constraint was anxiety. Two cautions. The account is retrospective and reconstructed, so it is a rationalisation as much as a record. And the framing is a structure for organising evidence, not a measurement model: do not assign weights to forces that were not measured. *Correct result:* four evidenced force accounts per change event studied, with the retrospective limitation stated.

**7. Distinguish an unmet need from an unarticulated preference, using the consequence test.** This is where needs analysis most often inflates. A preference is something a person would take if offered. **A need is unmet when its absence produces a consequence you can point at**: a workaround, an abandonment, a complaint, a substitute, an extra cost in money, time or effort, or a decision not made. Apply the test to every candidate. Then apply a second: **would the person notice if it were removed?** A preference disappears without a trace; a need leaves a hole. Candidates that fail both are reported as preferences, in a separate list, which is useful information and honestly labelled. Record prevalence per K2 §4.2 for every need retained, because "unmet need" is a phrase that inflates a single participant into a market opportunity faster than any other in research. *Correct result:* a needs list where every entry names its consequence evidence and its prevalence, and a separate preferences list.

**8. Separate stated barriers from operative ones.** People give reasons for not doing things, and the reasons are systematically biased towards the acceptable and the articulable. Price is the most over-stated barrier in research because it is a socially safe answer that closes the conversation; complexity, embarrassment, not knowing where to start, and not wanting to be seen as needing help are systematically under-stated. Four checks separate the two. **Does the behaviour change when the stated barrier is removed?** If price falls and adoption does not move, price was not operative. **Do people who face the barrier behave differently from people who do not?** **Does the stated barrier appear in the account of the moment, or only in the summary answer?** Barriers named in a narrative of what actually happened are more credible than barriers named in answer to "why not". **Do non-users and rejecters give the same barrier as users give retrospectively?** Report both the stated and the operative barrier, and say which is which. *Correct result:* a barrier table with the stated barrier, the evidence for whether it is operative, and the conclusion, per barrier.

**9. Find the needs that conflict within the same person.** Real people hold needs that pull against each other: wanting a decision made for them and wanting to stay in control, wanting speed and wanting reassurance, wanting privacy and wanting help. A tidy needs list has usually had these removed, and the removal destroys the most useful thing in the analysis, because the conflicts are where a proposition has to make a choice and where competitors are usually failing. Look for them explicitly: within a single participant's account, where does one stated need contradict another, and what did they do when both applied? The resolution behaviour is evidence about which need dominates in that situation. Report tensions as tensions, not as two independent needs. *Correct result:* a tensions section naming each conflict, the situations where it bites, and how participants resolved it.

**10. Assign confidence per need, per K3, and report the evidence with it.** Each need statement carries: the evidence type (workaround, consequence, substitute, direct statement), the prevalence with base, whether it was corroborated by behavioural data, whether it appeared across situations or only in one, and the confidence level. **A need evidenced only by what people said is at most moderate**, per K3 §3.4, because stated evidence is one step removed from the thing being claimed. A need evidenced by workarounds observed in several participants and corroborated by service records is a different kind of claim and should be stated with the strength it has earned. *Correct result:* a need statement register in which the confidence of each entry is visibly different where the evidence differs.

**11. Mark the judgement points and route them, per K5.** Whether a need is commercially material is a business judgement (§2.1). Whether the emotional and social readings hold in each market is a cultural one (§2.2). Whether a need the organisation cannot serve should be reported anyway is a strategic and professional judgement (§2.3), and the answer is usually yes, because an unservable need often defines where a competitor will enter. *Correct result:* markers at the point of the judgement, each naming what changes depending on the answer.

## 8. Analytical framework

**The five terms, fixed.**

| Term | Definition | Test | Typical error |
|---|---|---|---|
| **Want** | Something the person would like | Would they take it if offered? | Reported as a need because the word is stronger |
| **Need** | Something whose absence produces a consequence | Can you point at the consequence? | Asserted without consequence evidence |
| **Motivation** | The goal that makes the behaviour rational | Does it explain why this and not that? | Confused with a personality trait |
| **Barrier** | What prevents or slows the behaviour | Does removing it change behaviour? | Stated barrier taken as operative |
| **Job to be done** | The progress sought in a situation | Is it stated without naming a solution? | Written as a feature, so it names a solution |

**The elicitation ladder, and where to stop.**

    Attribute → Functional consequence → Personal consequence → Value
       specific        actionable            distinctive       universal
                          ↑                       ↑
                    usually the useful level for a decision

Climb from a specific instance. Stop at the last distinctive rung. Everything terminates in security, freedom, belonging or esteem if you keep going, and a framework that ends there could have been written before the study.

**The three dimensions of a job.**

| Dimension | Question | Evidence that supports it |
|---|---|---|
| Functional | What is the person trying to get done? | Behaviour, workarounds, task completion, substitutes |
| Emotional | How do they want to feel, or avoid feeling? | Emotional markers in accounts, avoidance behaviour, relief on resolution |
| Social | How do they want to be seen, and by whom? | Behaviour that is costly functionally and cheap socially, presence of others, concealment |

**The unmet need test.** A candidate must pass both.

1. **Consequence:** does its absence produce a workaround, an abandonment, a complaint, a substitute, or a cost paid?
2. **Hole:** would the person notice if it were removed?

Passing one only means it is a preference, and it is reported as one.

**Against the K2 chain**, a need statement is an Interpretation built on Findings about behaviour and accounts. It is never an Evidence-level claim, even when it uses a participant's words, and the output must make that visible per K2 §3.1.

## 9. Output format

**A. Situation inventory.** The circumstances in which the category arises, with their constraints.

**B. Need statement register**, one row per need.

| Need | Situation | Evidence type | Source and prevalence | Consequence evidence | Behavioural corroboration | Functional / emotional / social | Confidence |
|---|---|---|---|---|---|---|---|

**C. Job statements**, where the framing fits: situation plus progress sought, with the three dimensions, the evidence beneath each, and a stated assessment of how well the framing applies in this category.

**D. Laddering record.** Chains with participant-sourced rungs distinguished from analyst-inferred rungs, and the level chosen for reporting with the reason.

**E. Unmet needs**, each with the consequence evidence, the prevalence, and what is currently being used instead. Needs that failed the consequence test appear in a separate preferences list, labelled.

**F. Barrier table.**

| Stated barrier | Prevalence | Evidence it is operative | Evidence against | Conclusion: stated or operative |
|---|---|---|---|---|

**G. Tensions.** Conflicting needs within the same person, the situations where they bite, and how participants resolved them.

**H. Forces analysis**, where a moment of change was studied: push, pull, habit and anxiety, each evidenced separately, with the retrospective limitation stated.

**I. What could not be established**, per K3 §5.2: needs suspected but not evidenced, populations not covered (especially non-users), and the validation that would settle each.

**When the evidence is thin.** The needs list gets shorter, not more abstract. Raising the level of abstraction to cover weak evidence is the specific failure mode here, because an abstract need is unfalsifiable and therefore always supportable. A candidate with two mentions and no consequence evidence is a hypothesis with its validation named, per K3 §4.3. A category where the job framing does not fit gets motivations and occasions instead, with the reason stated. And where the sample contains no non-users, the barrier section says what it cannot cover rather than reporting user-recalled barriers as though they were the barriers that stopped anyone.

## 10. Quality checks

Run before anything is presented. These sit on top of K4 §8.

1. Does every need statement name the situation in which it arises?
2. Is any need resting solely on an answer to a direct question about needs?
3. Does every unmet need name its consequence evidence: a workaround, an abandonment, a complaint, a substitute or a cost paid?
4. Does every need carry its prevalence and base, per K2 §4.2?
5. Are analyst-inferred laddering rungs distinguished from participant-sourced ones?
6. Has any ladder been climbed to a universal value that would fit any category?
7. Do job statements avoid naming a solution, and is the fit of the framing to this category assessed rather than assumed?
8. Are emotional and social dimensions evidenced, or marked as inferred with a confidence level?
9. Is any inner state (feeling, fear, desire) presented as observed rather than interpreted?
10. Are stated barriers distinguished from operative ones, with the evidence for the distinction shown?
11. Does the sample include non-users or rejecters, and if not, is the limitation on barrier findings stated?
12. Are conflicting needs reported as tensions rather than as separate independent needs?
13. Does every need have a confidence level, and do needs with weaker evidence visibly carry lower confidence?
14. Has the abstraction level been chosen against the decision, and are the levels above and below shown?
15. Is any need present because a framework has a slot for it rather than because evidence supports it?
16. Are the K5 judgement points marked at the point of the judgement?

## 11. Common failure modes

| Failure | How to recognise it | How to prevent it |
|---|---|---|
| **The feature list called a needs framework** | Every "need" names a solution: "needs a faster app", "needs weekend opening" | Job statements must not name solutions; ladder up to consequence |
| **Asking people what they need** | The findings match what respondents said in answer to a direct question | Indirect elicitation: workarounds, abandonment, substitutes, consequences |
| **The universal need** | "Wants to feel in control", true of everyone in every category | Stop the ladder at the last distinctive rung; test whether it could be pasted into another study |
| **Preference inflated to unmet need** | An appealing idea with no evidence that its absence costs anyone anything | The consequence test and the hole test, both applied and recorded |
| **Prevalence omitted** | "Customers need X", from three participants out of twenty-two | Prevalence and base on every need, per K2 §4.2 |
| **Stated barrier taken as operative** | Price reported as the barrier, with no check on whether removing it changes behaviour | The four operative-barrier checks; non-user evidence |
| **Barriers from users only** | The sample is a customer list, and the barriers reported are the ones customers overcame | State the population limit; recruit rejecters where barriers matter |
| **The tidy needs list** | Six needs, no tensions, all mutually compatible | Look for conflicts explicitly; report tensions as tensions |
| **Inferred inner states as evidence** | "Participants felt patronised", with nothing showing it | Mark as interpretation with confidence, per K4 §3.4 |
| **Job framing forced onto a habitual category** | Strained job statements about routine, low-involvement consumption | Test the framing on two or three cases before committing; use motivations and occasions instead |
| **Jobs treated as a validated theory** | Job statements presented as measurements, with weights or scores attached | It is a framing for organising evidence. Do not weight what was not measured |
| **AI: fluent unevidenced needs** | A polished needs framework whose entries carry no source, no prevalence and no consequence | Every need traces to evidence with a base, per K4 §1. Fluency is the failure signal here |
| **AI: abstraction to cover thin evidence** | Needs stated so generally that no evidence could contradict them | Thin evidence shortens the list; it never raises the abstraction level |

## 12. AI guardrails

Skill-specific only. K4 applies in full and is not repeated here.

1. **Never state a need without its evidence type, its source and its prevalence.** A need without a base is an assertion in professional vocabulary.
2. **Never derive a needs framework from answers to direct questions about needs.** Report those answers as stated preferences and label them.
3. **Never call something an unmet need without consequence evidence**: a workaround, an abandonment, a complaint, a substitute or a cost paid.
4. **Never write a job statement that names a solution**, and never present a job statement as something a participant said. It is an interpretation and is labelled as one.
5. **Never present jobs-to-be-done as a validated theory**, and never attach weights, scores or sizes to jobs that were not measured.
6. **Never climb a ladder to a universal human value and report it as a finding.** If the statement would fit any category, it is not a result.
7. **Never assert an emotional or social dimension without evidence**, and never present an inferred feeling as observed.
8. **Never treat a stated barrier as operative** without the evidence that removing it changes behaviour, or that those facing it behave differently.
9. **Never report barrier findings from a user-only sample** without stating that the people whose barriers were operative are absent from it.
10. **Never remove a conflicting need to make the list coherent.** Tensions are reported as tensions.
11. **Never raise the level of abstraction to compensate for thin evidence.** A shorter list is the honest response, per K4 §1.

## 13. Best-practice principles

- **The words matter, and using them loosely is expensive.** A want, a need, a motivation, a barrier and a job lead to different decisions. Fix the definitions at the start of the project and hold them into the report.
- **People report solutions, not needs, and this is not a failure on their part.** Articulating a need requires introspective access nobody has on demand. Design the elicitation around behaviour and consequence instead.
- **A workaround is the strongest evidence of an unmet need there is**, because somebody spent real effort to compensate for something. Look for improvisation before you look for opinions.
- **Needs live in situations.** A need stated without its situation becomes an abstraction that fits everyone and guides nobody.
- **Stop the ladder while it is still distinctive.** Every chain in every category ends at security, freedom, belonging or esteem, and a framework that ends there could have been written before fieldwork.
- **The job is stable, the solution is not.** That is the real value of the framing, and it is why a jobs framework survives a technology change that would invalidate a feature-based one.
- **The social dimension explains the behaviour that looks irrational.** People will accept a worse functional outcome to avoid being seen as not knowing something, and the functional analysis alone will never find it.
- **Anxiety and habit usually beat push and pull.** Most adoption interventions are designed to increase attraction when the operative constraint was fear of the new or the cost of change.
- **The stated barrier is usually the acceptable one.** Price is the socially safest answer available and it is over-reported everywhere. Test whether it is operative rather than reporting it.
- **Contradictions inside one person are the most useful material in the study.** They are where a proposition must choose, and where competitors are usually failing.
- **An unservable need is still worth reporting.** It defines where somebody else will enter the category.
- **Thin evidence shortens the list; it never raises the abstraction.** An abstract need is unfalsifiable, which is exactly why it is the tempting way to hide weak evidence.

## 14. Worked example

*Fictional scenario, used to demonstrate method. The organisation, participants, figures and findings below are invented.*

**INPUT.** A regional health service wants to understand why uptake of its remote consultation service is far below forecast among adults aged over 60. Evidence: 24 depth interviews (P01 to P24) with people offered a remote appointment in the past three months, of whom 15 declined or did not attend and 9 used it; booking-system records showing where 3,400 journeys stopped; and a complaints log. The brief asks for "the unmet needs of older patients so we can improve the service".

**PROCESS.**

*Step 1, situations.* Three distinct situations emerge from the records and the interviews: a new symptom the person is worried about, a routine follow-up on a known condition, and a medication or administrative query. The needs in each are different enough that a single "needs of older patients" framework would have averaged them into uselessness. This finding alone reframed the brief.

*Step 2, indirect elicitation.* Asked directly, participants said they needed "someone to explain how it works" and "better internet", both solutions. The indirect evidence said something else. **Workarounds:** 9 of 24 described arranging for a family member to be present for the call, several taking time off work to do it. **Abandonment:** the booking records show 41% of remote bookings for the over-60 group stop at the identity-verification step, against 12% for under-40s. **Substitutes:** 6 participants described going to a pharmacy instead and asking there. **Consequence:** 4 described delaying care.

*Step 3, laddering, and the judgement call.* "I want someone with me" ladders to "so if they ask something I don't catch, it doesn't get missed" and then to "so I don't come away not knowing what I've been told". The next rung up is "so I feel in control of my health", which is true and would fit any health study in any country. **Judgement call:** stopping one rung lower is less impressive in a deck and vastly more actionable, because "coming away not knowing what you were told" is something a service can be designed to prevent and "feeling in control" is not. The lower rung is reported, with the higher one shown in the laddering record so the level can be revisited.

*Step 5, dimensions.* Functional: understand and retain what was said. Emotional: avoid the anxiety of uncertainty after the call. Social: **not be seen, by the clinician or by family, as someone who cannot cope on their own.** The social dimension is evidenced by 7 participants who described not asking the clinician to repeat something, and by 3 who arranged family presence but did not tell the service. It explains behaviour the functional analysis could not: people declining a service they could technically use.

*Step 7, the unmet need test.* "Wants a video option rather than telephone" is voiced by 5 participants and fails both tests: no consequence, no hole, nobody built a workaround for its absence. It goes in the preferences list. "Needs a durable record of what was said" passes: family-member workarounds, note-taking described by 6, two complaints in the log about instructions misremembered, and a substitute behaviour of going to the pharmacy to have it repeated.

*Step 8, barriers.* The stated barrier is technology confidence, given by 11 of 15 non-attenders. The operative barrier appears to be different: booking records show the drop-off concentrates at identity verification rather than at the call itself, and 5 of those who declined had completed comparable online tasks elsewhere. Technology confidence is reported as the stated barrier, with the evidence that verification friction and the social cost of asking for help are the operative ones, at moderate confidence because the interview base is small and the records cannot show intent.

*Step 9, tensions.* A clear conflict: wanting someone else present for reassurance, and not wanting to be seen as needing someone present. Participants resolved it by arranging family presence privately, which the service does not know about and therefore does not support. Reported as a tension, and it became the most acted-on finding.

**OUTPUT.** A situation inventory with three distinct situations; a need register with evidence type, prevalence and confidence per entry; job statements for two of the three situations and an explicit note that the framing fits the administrative-query situation poorly; a laddering record showing the reported level and the level above; unmet needs with consequence evidence, and a separate labelled preferences list; a barrier table separating stated from operative; a tensions section; and a **researcher review recommended** marker on whether a durable-record feature is deliverable within the service's information governance rules, which is a constraint the research cannot assess, per K5 §2.3.

## 15. Advanced usage

**Quantifying a qualitative needs framework.** Once needs are stated with their situations, they can be validated at scale: prevalence of each situation, prevalence of the consequence evidence, and importance within situation. Ask about the consequence rather than the need ("in the last six months, how often have you had to ask someone else to be present"), because a direct importance rating on an abstract need produces uniformly high scores and no discrimination. The output is sizeable needs, which is what a segmentation and a business case both require.

**Needs as segmentation bases.** Needs make the strongest basis variables available, per **09.01**, but only if they are measured as behavioural consequence items rather than as agreement statements about abstractions. A needs battery written as "it is important to me that..." produces an attitudinal battery with a needs label, and segments built on it will be scale-use artefacts as much as anything else.

**Non-category substitutes.** The most strategically useful question in this method is what people would use if the whole category vanished. The answer regularly identifies the real competitive set, which is frequently not the category's own competitors: a service competing with a family member, a spreadsheet, or doing nothing. Where the analysis has not identified any non-category substitute, it usually has not gone far enough up the ladder.

**Needs that the organisation cannot serve.** Report them, with the reason they cannot be served. They define the flank on which a new entrant will attack, and they are the input to the next planning cycle rather than to this one. Suppressing them because they are unactionable today is a version of cherry-picking under K4 §4.2.

**When the standard approach does not fit.** In habitual and low-involvement categories, job statements are strained and the useful frame is occasion plus motivation plus barrier, without the job layer. In highly regulated or professionally mediated categories, the person's job may be to satisfy someone else's requirement, and the job statement must name whose requirement. In B2B, the job belongs to a role in a process, and several roles have different and conflicting jobs in the same purchase, which is where the tensions analysis earns most of its value.

## 16. Skill chain

**Recommended previous skills:**
- **07.01 Thematic Analysis** and **07.03 Interview and Transcript Analysis.** Hand over coded material with prevalence, preserved contradictions and participant identifiers, which are the raw material of every need claim here.
- **09.03 Behavioural Profiling.** Hands over what people actually do and the occasion structure, which supplies the situations that needs must be attached to and the behavioural corroboration that raises confidence.
- **02.02 Discussion Guide Design** and **02.03 Interview Question Development.** Determine whether the fieldwork elicited behaviour and consequence, or only stated preferences, which sets the ceiling on what this skill can produce.
- **06.05 Customer Experience and Journey Analysis**, where abandonment and failure points have already been located.

**Recommended next skills:**
- **09.01 Audience Segmentation.** Takes needs dimensions as basis variables, measured as consequence rather than as agreement.
- **09.02 Persona Development.** Takes the goal, job, barrier and unmet-need elements with the evidence and confidence attached to each.
- **08.01 Finding to Insight Development.** Takes evidenced needs and explains the mechanism behind them, which this skill establishes but does not interpret.
- **08.05 Insight Prioritisation and Sizing.** Decides which needs are material, using the quantification described in Section 15.

**Runs well alongside:**
- **13.04 Bias Detection**, particularly where a solution existed before the research and the needs analysis may have been shaped to fit it.
- **13.05 Research Ethics and Consent Design**, wherever needs work reaches financial difficulty, health, capability or shame.
- **K3 §4.3**, for the hypothesis language that thinly evidenced candidate needs require, and **K5 §2.2** for the cultural reading of emotional and social dimensions.

---
A Yazi Supplied Skill and resource.