Your customer asks for another feature. What problem are they really trying to solve?

A latent need is a potentially valuable outcome that existing options do not sufficiently deliver, or that users do not yet express in a conventional product brief. It may be visible in frustration, elaborate workarounds, task failure, or an unspoken requirement for reassurance. It need not be unconscious in a clinical sense. In unfamiliar categories, people may simply be unable to describe a product that does not yet exist. Eric von Hippel's work on lead users illustrates how people facing future market conditions earlier than others can sometimes offer unusual insight into emerging needs [S03]. Urban and von Hippel subsequently studied how that information could contribute to actual industrial product concepts [S04].

However, a lead user is not automatically representative. Someone who modifies their own equipment may have a much higher tolerance for complexity, price or maintenance than the average buyer. The methodological answer is to use such cases for discovery, then test the identified problem and proposed solution with the intended market. A latent need becomes more credible when the difficulty recurs, there is an identifiable cost, the desired outcome makes sense, and a prototype can make a measurable difference. An engaging interview alone cannot establish that demand exists at commercial scale.

These rows illustrate hypotheses to investigate, not universal psychological translations. The same phrase can point to different unmet outcomes depending on the task and the individual.

Negative words associated with an existing solution can point towards a valuable opportunity. 'Leaks' may matter to users transporting food; 'unpredictable' to people using software; 'embarrassing' to someone operating a device in public. Yet the word alone does not tell designers what to build. A leakage concern might require a redesigned seal, clearer closure feedback, a better material, a user instruction, or simply quality assurance on the existing product. Each remedy has different cost and usability consequences.

Before drafting a feature, identify the associations that people spontaneously connect to the product category, the current alternatives and the desired outcome. Then look for the real occurrence behind each association. A newcomer might call a tool 'unreliable' because the feedback is unfamiliar. An experienced operator might report a genuine mechanical fault. Both expressions deserve attention; they do not justify the same product decision. Separate the researcher's inference from the user's words and from observed performance. The goal is neither to dismiss an emotional reaction nor to romanticise it as hidden truth. The goal is to turn the reaction into a falsifiable question about product value.

A customer's complaint can be produced by different types of friction. This framework prevents the team from assuming that better marketing or more functionality is always the answer.

Instead of opening a customer interview with 'Which innovation would you buy?', ask the participant to reconstruct a specific recent occasion. Where were you? What were you trying to get done? What was in your hands? Which option did you use and what happened next? Ask about the last occasion that went surprisingly well, too. Differences between success and failure often reveal more than a general satisfaction rating. Avoid suggesting that the feature your company has already developed is the solution.

Keep three columns in your research notes. The first contains observable events; the second preserves the participant's own language; the third records the research team's interpretation. 'I checked the lid twice' is an event. 'I was worried about leakage' is the participant's account. 'People need stronger closure feedback' is a hypothesis. The three belong together but must not become interchangeable. Griffin and Hauser explain the importance of moving systematically from customer input towards prioritised needs [S02]. For a product team, this separation also improves accountability: when a manager later questions a design choice, the researchers can trace the reasoning rather than replying that customers 'just felt' a certain way.

'Compact' may be a positive attribute on a crowded train and a limitation when someone needs to feed several people. 'Powerful' may sound reassuring in a technical sales pitch and exhausting during everyday use. Consumer choice research shows that preferences can be constructed within a decision context rather than being retrieved as fixed, context-free values [S07]. That insight changes how associations should be studied. The same person may connect a characteristic to different benefits depending on the task, the time pressure and the alternatives available.

Document the moment rather than relying on broad demographic labels. Is the product being used for the first time or as a habit? Alone or in front of colleagues? Under time pressure or with room to experiment? Ask whether 'unsafe' describes a real incident, an anticipated consequence or a general reputation. The answer determines what should be tested. An apparent market segment may actually be a situational pattern, so do not label customers too quickly. The earlier article about different responses within the same audience explores that distinction. Product research should add the next question: what job changes across those situations, and is there a design that can help without introducing new disadvantages?

Imagine an entirely fictional company called Lento, which makes reusable lunch containers. This is not a Neurofactor client, and the following names, experiences and designs are invented. Its team assumes the next product needs more compartments and greater capacity. In a hypothetical discovery exercise, Nora complains that her backpack sometimes ends up wet. Miles dislikes checking the closure when hurrying for a train. Amina cannot fit a bulky container into a shared fridge. Joel wants a lid he can open without strong gripping force. Their stated preferences appear different.

A possible common need is taking lunch along without unnecessary effort or uncertainty. It is an attractive hypothesis, not a measured finding. A real study would establish whether those experiences occur, their frequency and importance, and how different users respond. The team could then test several ideas: more reliable sealing, clear closure feedback, a smaller footprint, a more accessible grip, or fewer loose parts. The case illustrates the fundamental design choice. The company could build a feature customers already know how to request, or investigate the missing outcome that makes daily use inconvenient. Neither choice should be made solely on the basis of an elegant cluster diagram.

Suppose the container leaks only when used with particular contents or when packed in a way that the current instructions do not explain. The cause is different from a general mechanical weakness. Similarly, 'difficult to clean' may refer to a complicated shape, a faulty design, poor guidance or a specific cleaning context. Associative research helps surface the experience, but a product-performance investigation must establish what happens physically and why.

Proactive market orientation encourages firms to look beyond expressed requests [S01]. It does not exempt them from examining competing causes. For each interpretation, write down at least one alternative hypothesis and the observation that would disprove the preferred explanation. If a simple demonstration eliminates a misunderstanding and no genuine design disadvantage remains, a complex new mechanism may not be warranted. If the existing seal fails a repeatable technical test, clearer branding will not repair it. The challenge is to connect the value people want with the intervention that actually provides it. That is the difference between developing a solution because it sounds innovative and developing one because it survives testing.

A matrix helps the team preserve the chain of reasoning. It does not diagnose a market automatically. Each possible value claim should eventually be linked to observations and a verification method.

A negative expression is informative only in relation to other associations. 'Complex' accompanied by 'confusing', 'slow' and 'too many controls' suggests something different from 'complex' alongside 'capable', 'professional' and 'flexible'. Explore how meanings cluster and whether they refer to the same underlying experience. The Neurofactor knowledge-base article on association clusters explains the broader concept. Brand-association research likewise distinguishes properties such as strength, favourability, uniqueness and the origin of associations [S05][S06].

Avoid giving every visually striking cluster the status of a new buyer type. Check whether the pattern appears across participants, repeats at different times and predicts a meaningful difference in needs or responses. Preserve positive associations as well as negative ones. Removing every trace of complexity may damage a product that expert users value for its powerful controls. Likewise, adding reassuring cues may weaken a distinctive simplicity. The product team needs a map of what is currently valuable, what is missing and which trade-offs people will accept. Without those distinctions, a beautifully designed association map can become an expensive way of justifying an existing product roadmap.

'Our innovative closure will delight customers' is not a research hypothesis. 'A visible closed-state indicator will reduce unnecessary checking among commuters without increasing closure time or errors' is much closer. It names a situation, intervention, expected outcome and possible side effect. Define before testing how much improvement would matter, how failure will be recognised and what alternative mechanism should be examined.

A practical hypothesis register includes the association, observed friction, desired outcome, candidate prototype, primary measure, possible adverse consequence and decision rule. The primary measure should normally reflect the promised task benefit. An association rating can help explain why an option is attractive, but cannot replace functional success. Research on innovation spans need discovery, consumer response, product development and market entry, with different tools suited to different questions [S13]. Do not let a polished concept test answer questions it was never designed to investigate. A disciplined team allows the possibility that the existing product is sufficient, or that a modest manufacturing fix will achieve more than an expensive new feature.

The hypothetical alternatives below are deliberately different. They test competing ways to deliver the same broad benefit rather than competing expressions of the same marketing claim.

A company might discover a valuable missing outcome and still fail to make the product commercially viable. A buyer saying 'too expensive' may mean the difference is unclear, the alternative already works or the risk of disappointment is too high. But the objection may also be literal: the product exceeds what the person can or wishes to spend. Association analysis must not psychologise away ordinary budget constraints.

Separate the importance of the unmet outcome from proof that a new design delivers it and from willingness and ability to pay. Someone may genuinely value a reliably sealed container while having no interest in a premium price. The answer could be lower-cost engineering, a different version, a changed service model or a decision not to develop. Product strategy is not about maximising perceived desirability in isolation. It is about finding a feasible relationship between the value users recognise, the evidence behind it, the purchase situation and the cost of delivery. The same discipline protects product teams from persuasive but financially misleading innovation stories.

Return to the fictional container. Nora travels with electronics and interprets 'leak-proof' as avoiding damage. Miles prepares food before leaving home and hears it as freedom from repeatedly checking the lid. Amina needs something that fits in a small shared refrigerator and values compact storage. Joel needs an opening action that is manageable without high gripping force. These people are invented, but their situations demonstrate why a single product attribute may support different user outcomes.

Do not turn them into four personality categories. Actual research would observe situations, allow participants to disagree and check whether patterns repeat in new users. Some contexts may be addressed by one shared product promise: dependable carrying with fewer unnecessary actions. Other requirements may conflict, such as maximum resistance to opening versus ease of opening. That can justify different designs or a conscious product choice. The related article on how the same need can attract and repel consumers explores such tension. The innovation question is whether the company can serve more than one outcome without weakening the task the product exists to perform.

Step 1 - define the task. Identify what the customer is trying to accomplish, not just the category they buy. Step 2 - inventory alternatives. Include workarounds, postponed purchases and non-use. Step 3 - observe real occasions. Note the context, actions, effort and errors. Step 4 - elicit spontaneous associations. Avoid leading respondents towards a proposed design. Step 5 - separate meaning from cause. Fear of failure is distinct from observed mechanical failure.

Step 6 - structure the need hypotheses. Seek recurring patterns but retain outliers and counterexamples. Step 7 - develop competing concepts. Give each one a clear mechanism and a possible downside. Step 8 - test under realistic conditions. Compare against the current alternative on task success, experience and unintended consequences. Step 9 - make a decision. Improve, reject, narrow the market, or investigate further using predefined evidence thresholds. This is a practical editorial research protocol, not a proprietary nine-stage method claimed to have been validated through Neurofactor experiments. Keep a written record of failed hypotheses, too. They often prevent the next team from making the same expensive mistake.

A prototype image tells people what they expect. Hands-on use reveals what happens. Repeated use shows whether the improvement holds once novelty disappears. These stages may produce different associations. Measure expectations and spontaneous meanings before first contact. Observe task performance, confidence, errors and workarounds during use. Then ask whether the anticipated benefit was delivered and what the person would choose next time. If possible, include a later follow-up rather than relying on a favourable first impression.

Van Osselaer and Janiszewski examine different processes through which people learn feature-benefit associations [S14]. Lemon and Verhoef place customer experiences within an evolving series of touchpoints rather than a single isolated rating [S15]. Neither study proves that a given container design will succeed. Their relevance is methodological: meaning can change as someone learns what an option actually does. The prototype that sounds impressive on a concept board might disappoint in a real bag. A strange-looking solution might earn trust after a week of reliable use. Innovation decisions should respect that sequence.

No single measure can establish unmet need, psychological meaning, technical feasibility and market demand all at once. Use methods in a deliberate combination.

Suppose customers who love a product frequently say 'secure'. Maybe the product makes them feel secure. Or perhaps generally enthusiastic customers use positive language across every topic. Their experience, familiarity or the specific product version could explain both the association and the positive rating. An association cluster can help identify promising explanatory variables without showing which direction the influence runs.

When possible, randomise prototype order or exposure, keep competing factors constant and decide in advance what the primary outcome will be. If the question concerns a lid mechanism, hold packaging, pricing and instructional language stable across versions. Confirm basic functional safety before exposing participants to realistic use. Capture the associations before and after the task and inspect negative effects alongside improvements. An indicator that looks reassuring but causes more closure errors is not a success. The Neurofactor knowledge-base article on measuring associations discusses why an association claim needs more than a collection of evocative words. Causal language belongs only to designs that actually support causal inference, and any claim must match what was measured.

Research workshops often end with walls of language: ease, freedom, reliability, elegance, peace of mind, value, simplicity. The product department still needs to decide what to build. A useful decision statement sounds like this: 'We will prioritise clearer closure feedback because unnecessary checking recurred in observed tasks and this prototype reduced those checks without introducing errors.' That statement requires real data. Before testing, it is a hypothesis and should be labelled as such.

A decision record can use five questions. Is the problem recurring and important? Does the prototype outperform the existing option on the actual task? Is its benefit understandable and credible? What disadvantages appear for other users or contexts? Is the result feasible, safe, affordable and socially responsible? Those final conditions are not side notes. A product that increases waste, prevents accessible use or creates a maintenance problem can have attractive associations while being a poor innovation. Association research should illuminate the whole trade-off, not decorate the favourite answer.

Imagine Lento develops a container with four additional compartments because customers requested 'more possibilities'. The launch visuals look impressive. In a hypothetical user session, participants leave compartments empty, dislike washing the extra pieces and continue checking the lid before putting it in a bag. These observations are fictional, included only to show how an apparently sensible product enhancement might miss the core difficulty.

A second hypothetical design offers clearer closure feedback and fewer removable parts. It might reduce checking and improve task ease, but it must still pass mechanical sealing, cleaning and durability tests. Only actual comparative evidence could establish that it performs better. The distinction has marketing consequences too. The first design leads with a list of extra components; the second could, if supported by testing, promise confidence with less effort. Our related article on credible brand positioning from associations explains why distinctive claims must connect to proof rather than appealing copy alone. Product innovation and positioning are linked, but they are not the same analytical step.

Treat this table as a claim-control tool. It makes the route from hypothesis to design and evidence traceable, and it prevents speculative associations from becoming factual marketing statements.

Neurofactor's contribution can be to clarify the meaning people connect to a product alternative, a friction point or a proposed benefit. Interview work, structured association research and, where methodologically appropriate, indirect measures may help identify relevant questions. Those findings can then inform prototype comparisons and the language used to explain them. The broader article on associations in marketing explains why meanings matter for decisions. But this is only one part of development. Context observation, design engineering, price research and product-safety tests must remain part of the process.

Neither EEG nor an implicit measure can directly tell a company which feature will sell, how long a seal will last, or whether a prototype meets regulations. Neurofactor-specific research outcomes must be assessed within their real sample, design and publication permissions. This article therefore makes no claims about actual client results and provides no invented effect sizes. The scientific literature supplies tested ideas and methodological cautions, not automatic forecasts for this fictional example. Stronger products are built when the experience researchers and engineers can challenge one another's assumptions with relevant evidence.

Build three decision gates into product development. Discovery gate: is there a recurring problem with a meaningful cost, or merely a persuasive anecdote? Concept gate: does the proposed solution change the right expectation and avoid unacceptable side effects? Use gate: does it improve the real task after hands-on or repeated use, at a sustainable cost and with acceptable safety and accessibility? At each gate, ask a colleague outside the project to offer the strongest alternative explanation.

Define stop rules as well as go rules. Stop developing when the problem is too rare, when the new solution makes everyday tasks harder, or when benefits can be achieved by repairing the existing option. Revise the positioning when the product works but people misunderstand it. Return to engineering when the association score improves but the functional performance worsens. Ajzen's theory of planned behaviour distinguishes intentions and behaviour and gives a role to perceived behavioural control [S16]. It is therefore wise not to treat high reported purchase intention as sufficient evidence of repeat use, adoption or commercial success. The decision should be anchored in the outcome the innovation promises to improve.

The most valuable innovation idea may not begin with somebody asking for something new. It may begin with a routine detour they would be relieved to stop taking: checking a lid twice, carrying a spare bag or memorising a complicated sequence. Negative associations can draw attention to the problem, while careful observation reveals its actual cost. Neither is enough by itself. The team must test alternative causes, compare different solutions and establish that the missing outcome can genuinely be delivered.

An unmet product need is not proven by an evocative cluster of words. It becomes credible when a relevant outcome is missing and a better way of achieving it survives real use. Companies that understand that distinction can stop adding features simply to look innovative and start designing products that remove the right friction.