Your customer calls themselves loyal. Until another brand is on the shelf.

AI summary: Loyal customers can switch brands because positive associations do not equal repeated purchases or exclusive preference. The situation, availability, prices, changing needs, competition, habit and mental accessibility may all influence the next choice. Examine separately what customers spontaneously connect to a brand, which alternatives they consider and what they actually buy. Combine association research with behavioural evidence, test a specific improvement and avoid claiming that EEG or a loyalty score can predict individual future purchases.

Imagine someone who has bought the same coffee for two years, follows the brand online and speaks warmly about its flavour. At work, however, the vending machine stocks a different brand. In the supermarket, their usual bag is unavailable. When visiting friends, they discover a different roast that suits a particular occasion. None of these situations requires the customer to stop liking the original brand. They show that brand preference does not grant a company ownership of future transactions.

Marketing departments sometimes make the relationship too absolute. High satisfaction becomes a promised retention rate. A positive brand-association score is interpreted as guaranteed share of wallet. Someone signed up to the newsletter is labelled loyal even when they follow several brands primarily for special offers. The reverse occurs, too: a customer repeatedly buys a product because it is convenient to find while feeling little attachment to the name on the packaging. These distinctions matter when allocating budgets to campaigns, packaging, rewards and distribution.

Dick and Basu's conceptual framework distinguishes relative attitude from repeat patronage, while recognising the roles of situational conditions and social norms. [1] Oliver discusses why satisfaction does not always become loyalty and why loyalty may depend on more than a satisfying experience. [2] These are theoretical foundations for investigation, not personality tests that reveal an individual's next purchase.

One explanation may be the changing meaning of the buying occasion. 'My reliable coffee for early mornings' does not necessarily compete with 'something special for guests'. 'My usual sportswear brand' competes, during an urgent purchase, with 'available today'. 'I like using this software' may give way to 'it must integrate with our existing systems' in a team procurement decision. The original brand remains positively evaluated, but the job to be done has changed.

Research therefore needs two questions. Which favourable associations are strong and accessible enough to bring the brand to mind on relevant occasions? And which additional meanings or practical constraints then lead people to choose another provider? Understanding both is more useful than treating every purchase outside the brand as a betrayal.

These eight possibilities form a diagnostic map, not a set of established causes for every customer. 'Expensive' may indicate premium quality to one person and unaffordability to another. Researchers must not automatically assign a positive or negative score without examining what a term refers to and what consequence the customer draws from it.

There is also no magic threshold above which brand loyalty becomes permanent. A category purchased once a year differs from a weekly grocery category. A small brand has a different penetration level from a market leader. Research on the double jeopardy pattern describes how smaller brands tend to have both fewer buyers and buyers who purchase them less frequently. [7] That is a market-level pattern with exceptions, not proof that an individual customer defected because of weak brand associations.

The real question is not how many loyalty-related words people can attach to a brand. It is which expectations matter at the next purchase, whether the brand is recognised as a relevant option and whether it is available and suitable. An exceptionally well-liked product can still fail to be purchased if it is not sold where the buyer shops.

An account manager asks whether the customer is satisfied. A brand researcher asks which brand is most appealing. A data analyst studies transactions. Each observes a different part of the relationship. Problems begin when all three findings are combined into one supposedly complete 'loyalty score'. A high recommendation intention can exist alongside low purchasing frequency. Frequent purchasing can result from convenience, contractual conditions or weak alternatives. Strong affection can survive an occasional purchase elsewhere.

Chaudhuri and Holbrook examine purchase loyalty and attitudinal loyalty as separate variables in a model of brand trust, brand affect and performance. [3] Their analysis considers brands at an aggregate level; it does not establish a rule that predicts individual buying behaviour from brain data. Its practical value is in stopping the convenient leap from positive association to guaranteed sales.

Seiders and colleagues examine conditions under which satisfaction is associated with higher purchasing in a retail setting. Their research illustrates why the satisfaction-to-behaviour relationship is conditional. [4] Gustafsson and colleagues study how satisfaction, different kinds of commitment and triggers relate to retention in telecommunications. [5] These studies offer important questions, but their findings should not be transferred to a specific brand as promises of conversion lift.

Set three definitions before analysing loyalty. Attitude: how much value, confidence or connection does the customer feel? Behaviour: how often does the person choose the brand within a specified period and category? Context: when was the brand a genuine option, and which alternatives were available? Without the third dimension, comparisons of the first two often produce misleading contradictions.

Consider a buyer who admires a brand but purchases a substitute once because their preferred variant is out of stock. That differs from a customer who permanently switches after repeated service failures. Both differ from someone who buys a luxury product once a year because that is the natural replacement frequency. A useful strategy starts with these distinctions instead of trying to make one indicator explain everything.

A brand is never selected in a vacuum. The buyer has a reason to purchase, a certain amount of time, available information, a shopping channel and sometimes other people to consult. The first purchase may be driven by curiosity. The next may be about saving time. Packaging could matter greatly when buying a gift. Availability or compatibility may dominate when buying a routine replacement. This does not mean the brain learns the brand again from scratch every time. Rather, different situations can activate different relevant expectations and associations.

Do not rely solely on the question 'Why did you not choose us?' That phrasing demands a retrospective explanation even for a largely habitual or incidental choice. Ask about the exact situation. Was the brand actually stocked? Which option did the person first notice? What problem needed solving that day? Which alternatives were genuinely considered? How did price, accessibility, delivery and terms compare? And how did the customer evaluate the original brand apart from that transaction?

An adverse event can change existing associations: an unresolved complaint might weaken 'dependable'. Yet a neutral change can also redirect the choice: a different household, store, procurement policy, subscription model or usage requirement. A model based only on brand reputation misses those changes in the choice environment.

Wanting variety does not automatically mean disloyalty. Customers may intentionally use different brands for different purposes: one coffee at home, another at the office and an unusual roast when friends visit. Brands often exist within a repertoire rather than an exclusive relationship. Evidence on double jeopardy and repeat purchasing can help marketers develop more realistic ambitions. [7] No single consumer should, however, be mechanically classified from a market-wide pattern.

A useful association question is: for which situations should our brand come naturally to mind, and where is it not relevant? That supports more precise choices than simply trying to increase general positivity. A brand associated with premium gifts does not necessarily need to win everyday bargain purchases. A complex business service provider need not pretend it is the ideal solution for every quick administrative task. Strong positioning can include a clear understanding of its limits.

This entire case is fictional. Zuidlicht Coffee, the shoppers, store observations, concepts and proposed interventions do not come from a measured Neurofactor client project. No purchasing effects or research outcomes have been observed.

Zuidlicht Coffee is an imaginary roaster with recognisable dark-blue packaging and an everyday range. In fictional customer conversations, people value the flavour and consistent product experience. The marketing team sees enthusiastic social engagement and favourable newsletter responses. Yet the sales team notices that many customers do not choose Zuidlicht whenever they buy coffee. Their first instinct is a loyalty initiative: more points, a larger discount for the second order and a stronger 'thank you for being loyal' message.

That response may begin in the wrong place. In this thought experiment, three buying contexts matter. Home mornings: the buyer wants a familiar taste without searching. On the move: the purchase depends partly on what a supermarket or station sells. Entertaining guests: the buyer wants something distinctive and is more open to variety. One person can have all three needs. They are different purchase occasions, not fixed psychological consumer types.

One hypothetical participant says: 'I really like Zuidlicht, but I happened to see another bag.' Another explains that they deliberately try new flavours for special occasions. A third loves the packaging but is unsure which variant works with a recently purchased brewing device. If real customers said such things, the possibilities would include distribution, exploration and product fit. Without actual research, none of these should be presented as an established market finding.

The team develops three untested concepts. A - recognition and availability: improve visibility within relevant channels, within actual distribution capabilities. B - suitability and evidence: explain brewing methods and usage occasions on packaging and product pages, using accurate information. C - transparent value: show the real benefits, total costs and terms of repeat orders or subscriptions, with simple cancellation. No version is an assumed winner.

The correct intervention might still be unrelated to these concepts. If Zuidlicht is not stocked in the buyer's store, more brand persuasion does not fill the shelf. A discount may pull a future purchase forward without permanently changing preference. Better information about a brewing method may be useful but irrelevant to someone buying a different flavour solely for variety. The results need to determine the solution, not the team's preferred campaign.

A proper study would give participants realistic tasks covering different occasions, for example coffee for a busy working week followed by coffee to serve guests. Observe which brands they spontaneously consider, what details they need and which factors change their choice. If the available traffic, sample and sales data allow it, evaluate interventions against relevant actual purchases over an appropriate period. Report situations in which the original brand quite reasonably loses.

The strategic lesson is that brand meaning must be related to a concrete buying occasion. If Zuidlicht is familiar at breakfast but irrelevant to hospitality, broadening that association may be a testable opportunity. If the brand cannot be found, the response is partly a channel decision. If customers do see it but cannot identify the right variant, it is an information or design problem. Three different issues should not be hidden under one vaguely defined loyalty campaign.

A competent researcher does not treat an observation as its cause. Falling sales among existing customers could reflect weakened associations, but the assortment or the category may also have changed. Even describing somebody as 'lost' requires care. A person who has not purchased shampoo for six months may still have enough at home. A business buying expensive machinery every three years does not belong in a monthly consumer-style churn report.

Use three layers of evidence. First, the feasible choices: what category purchases and alternatives were available? Second, brand meaning: which associations emerged spontaneously in relevant situations? Third, actual behaviour: what did people choose, and was that an appropriate decision for their needs? Without these layers, a neat loyalty dashboard may confuse preference, availability and random variation.

Check what your data cannot observe. Your own CRM records purchases from your company, but not necessarily whether the customer has stopped buying the category or increased spending elsewhere. Wider panel studies require appropriate permission, proportionality and a clear research purpose. A missing transaction does not identify the competing brand or reveal the customer's thoughts. The temptation to infer 'customer switched to competitor X' from a blank row should be resisted.

This is a translation tool between marketing strategy and research. Rather than starting with an advertisement, ask which conclusion customers still cannot justifiably reach. If the goal is 'easy to find', the distribution must support it. If it is 'sustainable', suitable product evidence is necessary. Marketing can communicate genuine facts more effectively; it must not invent those facts.

Nor can an association be guaranteed by repeating a tagline. A brand claiming to be 'perfect for every occasion' while serving a very narrow use case may lose credibility. There is a difference between what the brand would like to mean and what customers need to decide. The relevant situation should define the test. A strong brand image without relevance to the actual task is not automatically a competitive advantage.

Sometimes the best choice is not to pursue every customer. Improving availability, developing a relevant variant, clarifying use cases or becoming more easily considered for one extra shopping occasion can be more useful than a broad promise to 'increase loyalty'. Each option should be evaluated through the effect it plausibly changes, rather than through the appeal of the marketing language.

Step 1 - Define loyalty before measuring it. Decide whether the research concerns attitudinal preference, repeat purchases, retention, category penetration or share of wallet. A subscription that continues by contract is not the same as a voluntary new choice. Specify the observation period, category and number of genuine repurchase opportunities. Also define an appropriate outcome for the customer. More automatic payments are not necessarily better value.

Step 2 - Audit the real choice environment. Examine inventory, distribution, range, shipping, ease of access, prices and temporary promotions before searching for emotional causes. Check whether a favourite variant disappeared or whether the brand was even available in the chosen channel. Group differences may arise from channel access rather than permanent psychological differences. Record which purchase occasions you can actually observe.

Step 3 - Investigate usage occasions. Ask participants to reconstruct a specific recent purchase and a realistic upcoming shopping task. What was the product needed for? Was there time pressure? Did anyone else influence the choice? What alternatives were seriously considered? Avoid asking only why they did not select you. A routine decision can acquire an elaborate rational explanation after the fact, which need not represent the original process.

Step 4 - Map the actual associations. Examine spontaneously mentioned words, situations, expectations and feelings concerning the focal brand and relevant alternatives. Distinguish brand, retailer, product and variant when appropriate. A retailer may have poor service while the customer's view of the manufacturer remains positive. Identify plausible missing, desirable and negative associations, but label them as hypotheses until tested.

Step 5 - Compare contexts without stereotyping. Look at genuine needs, new versus experienced customers, relevant buying channels and category behaviour. Check sample sizes and whether a word has the same meaning across groups. Where justified, carefully designed implicit tasks can add relative association measures. Reaction times are not universal loyalty scores. Explain instrument validity, reliability and possible alternative explanations.

Step 6 - Observe realistic choices. Give people a task with credible alternatives and permission to choose a rival, postpone or not purchase. Watch which brand is found, whether the right variant is recognised and what evidence is consulted. Eye tracking can help investigate visibility but does not establish affection or loyalty. Pair attention data with comprehension and choice results, remembering that a laboratory task differs from a busy real store.

Step 7 - State one falsifiable intervention hypothesis. For example: 'If appropriate coffee variants become easier to identify for everyday brewing tasks, returning customers may select a suitable Zuidlicht option more reliably.' Specify the occasion, proposed mechanism, evidence, outcome and possible harms. Avoid changing discounts, package design and distribution simultaneously if you later want to know which element caused the difference.

Step 8 - Test real purchasing and choice outcomes. Prefer a prespecified, adequately powered experiment with comparable groups when practical. Measure relevant orders and, when possible, appropriate purchases, not only brand recall and clicks. Account for stock, promotion, season, purchase frequency and category fluctuations. Report denominators and uncertainty rather than just dramatic relative changes. If the evaluation is observational, explain why causal claims remain limited.

Step 9 - Examine lasting value. Investigate whether behaviour persists when a promotion ends, whether margins remain healthy and whether product use matches the promise. Compare intended changes in association with purchase behaviour in the same relevant settings. Do not remove inconvenient results from the report. Assign ownership of follow-up research, customer experience, operational issues and product changes.

The protocol is not a mandatory laboratory programme. A small brand with a demonstrable stock problem should begin with operations. A brand with good distribution data but unclear meaning may benefit from qualitative and association research. Choose the method because it distinguishes a relevant hypothesis, not because sophisticated measurement looks impressive. Whenever customer identities and purchase records are linked, follow applicable privacy and consent requirements.

1. Design around occasions, not a mythical perfectly loyal buyer. Identify the situation in which the product helps and which outcome the buyer values. Communicate why the relevant variant fits. Accept that another product may be a better choice in other situations. A specific use case is more actionable than trying to be everyone's permanent favourite.

2. Treat availability as part of brand strategy. A positive impression cannot solve a stockout. Check shelves, searches, product filters, channel coverage and delivery. Distinguish operational work from communications. An advertising campaign can help customers find a genuinely available product; it should not imply distribution the company does not have.

3. Make variant differences credible. Explain which features genuinely matter to the customer's task. Avoid marketing every product as universally superior. Show relevant evidence, conditions and limitations. Test whether people can make a suitable selection, not only whether they find the copy attractive.

4. Allow customers to seek variety. A consumer buying another brand occasionally has not committed an offence against your brand. Position products for situations in which they are genuinely useful. Consider range or complementary needs without insisting on exclusive loyalty. Test whether broader relevance fits the brand and delivers value.

5. Support trust through actual performance. A slogan cannot repair poor delivery, broken promises or evasive service. Make truthful evidence and policies accessible. Real reviews are useful when their relevance and provenance are clear. Invented ratings, testimonials or certifications are not legitimate instruments for generating loyalty.

6. Explain value without funding every transaction. Discounts can affect short-term choices and attract promotion-driven purchases. State true prices and relevant benefits. Do not equate discount redemption with stronger emotional loyalty. Measure full margin and purchasing behaviour after incentives end.

7. Communicate when a reminder is genuinely useful. A replenishment reminder can serve a real need if timing is appropriate and the customer has agreed to receive it. An indiscriminate weekly message may create irritation. Respect actual consumption patterns and make opt-outs straightforward. Permission is more valuable than forced attention.

8. Keep the brand promise consistent with experience. A brand claiming simplicity should not have a needlessly complex checkout. A brand associated with helpful service should offer real help at the moments it matters. Identify where experience reinforces the promised meaning and where it undermines it. An unpleasant return process can outweigh a polished campaign.

9. Investigate better competitive solutions. Competitors may actually meet a changing need better. Study that possibility without attacking their reputation. If your product or range has fallen behind, the solution may be product improvement. Repeating the same brand promise cannot manufacture suitability.

10. Measure the relationship rather than forcing retention. A suitable, profitable repeat purchase is more useful than an order extracted through misleading terms. Track returns, complaints, service costs and customer outcomes. Let people leave subscriptions easily. A brand retaining users only through barriers has not necessarily earned positive commitment.

These decisions are directions for hypothesis-driven change, not guaranteed conversion tactics. They belong on a roadmap only when the evidence suggests the relevant mechanism is plausible. Results from one category may not carry over to another with different usage and purchasing conditions.

A customer may rate a brand very favourably when prompted yet fail to think of it spontaneously at a shopping occasion. That is why brand research should not rely entirely on presenting a list of brand names and asking for an opinion. Ask which names people independently retrieve when thinking about specific tasks, needs and occasions. Investigate what situations are linked to the brand in their own language. A brand need not be salient everywhere to be genuinely relevant in its category.

Separate the first brand considered from the brand eventually purchased. Someone may think of you but choose another product because of price or availability. Another person may buy your product without any emotional commitment because it is in a convenient location. Calling both events 'loyalty' hides what the organisation can improve.

Useful reports therefore distinguish spontaneous consideration, aided recognition, credibility and observed choice. The goal is not a complicated dashboard but clearer decisions. Weak recognition may call for consistent brand identifiers. Missing stock calls for operational action. An inadequate fit between product and category needs may require new positioning or an adjusted offer.

The double jeopardy findings of Ehrenberg, Goodhardt and Barwise provide valuable context. [7] Brands operate within markets with different penetration and purchasing frequency, so loyalty percentages cannot be interpreted without the structure of the category. However, a market-level pattern is not proof that any one campaign designed to increase mental accessibility will lift sales. That needs appropriate testing.

Mental accessibility is also an ethical and scientific concept, not a claim that marketers can read thoughts or programme people to buy. Research can examine associations, alternatives and choices under defined conditions. A sophisticated lab measurement does not automatically become a valid individual loyalty diagnosis. When the simpler behavioural evidence answers the question, it should take priority.

Explicit association research explores the terms, expectations, occasions and feelings customers spontaneously link to brands. Open questions can reveal ambivalence rather than an artificially neat positive score. 'Familiar' might be comforting and boring; 'sustainable' might imply credibility and high prices. Clarify whether a judgement concerns the product brand, retailer, product variant or context. Results describe meanings within the studied sample, not guaranteed future purchases.

Implicit association tasks may provide complementary information about relative links between categories and concepts when the measure is properly designed. Reaction-time differences do not directly show who is loyal or who will buy next week. Task construction, reliability, construct validity, stimuli and interpretation must be documented. Use such methods only when they answer a question the available behavioural evidence does not already resolve.

Eye tracking can help establish how customers visually navigate a shelf, pack or website. Did they see the brand? Was an important feature missed? Was the appropriate variant easy to identify? Gaze fixation is not a measurement of trust or preference: prolonged attention may reflect uncertainty as easily as interest. Combine gaze results with task outcomes and comprehension. Consider the limitations of a controlled screen-based simulation compared with real-world shopping.

EEG records electrical brain activity under defined conditions. With a suitable time-locked stimulus and experimental design, it can support investigation of neural responses. It does not provide a universal direct scale of brand loyalty, trust or 'customer will purchase this again'. A signal may be affected by attention, task difficulty, artefacts and stimulus context. Interpretation requires validation and corroborating behaviour. For many loyalty questions, properly analysed transactions and qualitative choice research should come first.

Behaviour and economic context complete the picture. Retention, repurchase intervals, spending share, category activity and distribution capture distinct outcomes. Verhoef's CRM research considers retention and customer share development within an explicitly relational setting. [6] Such work does not justify converting a neural score into a sales forecast. Research is most useful when different evidence streams can confirm or challenge the same practical explanation.

Privacy cannot be an afterthought. Linking identifiable shopping histories, locations, reaction times and physiological measures raises questions about necessity, legal basis, transparency, access and retention. Aggregate findings will often suffice. Do not construct individual 'disloyalty profiles' from brain measures or treat a customer who explores alternatives as doing something wrong.

Loyalty programmes can provide value, but membership is not proof of exclusive preference. Shoppers may participate in several programmes and select retailers according to location or offers. Meyer-Waarden studied relationships between reward programmes, customer lifetime duration and share of wallet in grocery retail. [8] The findings are useful while remaining dependent on category, competition and programme design. They do not mean any points scheme automatically strengthens the desired brand meaning.

Verhoef's analysis in financial services shows that relationship instruments and perceptions are associated differently with retention and changes in customer share. [6] This supports measuring more than one outcome, not sending more messages to every customer by default. A financial-services relationship does not function identically to an everyday product sold through several independent shops.

Recommendation intention is another distinct measure. Net Promoter Score records what respondents say about likelihood to recommend in that study context. Keiningham and colleagues critically examined claims that NPS is uniquely superior at predicting business growth and did not find those universal superiority claims supported. [9] NPS may form one part of the research, but cannot replace a purchase history or explain a particular choice by itself.

Kumar and Shah examine loyalty in connection with sustainable customer profitability and long-term value. [10] This matters because repeat orders purchased with heavy discounts can create a misleading success story. A promotion that shifts future demand into the current month might increase transactions during the campaign without creating lasting new value. Margin, displacement and customer benefit must be included in the evaluation.

The conclusion is not that loyalty programmes are bad. It is that their intended mechanism, consumer benefit and measurable effect need to be clear. A useful programme can make relevant shopping easier and give participants legitimate benefits. Problems start when participation is presented as evidence of deep emotional attachment while actual alternatives and customer attitudes have never been examined.

A brand can remain well liked while sales drop for plain operational reasons. A product is out of stock, a size disappears, quality deteriorates, a competing product is materially more suitable or the customer's preferred shop no longer stocks the brand. Search filters may miscategorise an item. A business may change its procurement rules. Those problems need factual diagnosis and, often, operational change instead of a new branding campaign.

Changing needs are another limit. Someone who has changed diet or moved to a different transport mode may no longer need the original category. Targeting such a person with repeated generic retention discounts can feel intrusive and be wasteful. Some relationships end naturally, without anyone failing. A decline in purchase frequency is not always evidence of defective marketing.

Measurement limits matter too. A small self-selected newsletter survey is not automatically representative of the customer base. A simultaneous stock change, price change and design change cannot isolate the causal effect of the new design. Rarely purchased products are especially vulnerable to misleading short observation windows. Do not replace that uncertainty with a loyalty figure reported to a misleading number of decimal places.

Positive associations also require ethical boundaries. Claims concerning health, safety, sustainability or environmental impact need appropriate evidence. Colour, design and suggestion must not strengthen claims the product cannot support. A loyal customer deserves accurate information, not fabricated reassurance. Strong affection is not permission to overpromise.

Choose the primary outcome before testing. For a grocery brand, it may be an appropriate repeat purchase within an economically relevant period, with penetration and distribution as supporting measures. In subscription services, distinguish voluntary retention from payments that continue because cancelling is difficult. In B2B markets, procurement cycles may be so long that monthly repeat-purchase rates make little sense. The metric has to match the underlying buying decision.

Use a clear measurement sequence. Visibility: can the brand be found in the relevant channel? Meaning: are the intended, credible associations accessible in that situation? Consideration: is it among real alternatives? Choice: is it bought when suitable and available? Experience: does the product fulfil its promise? Repeat: is it chosen when a new relevant occasion arises? These are different questions, not an automatically causal funnel.

Report absolute as well as relative changes. A rise from two purchases to three is 50% in relative terms but just one additional purchase. With small numbers, chance can dominate. State the minimum evidence needed, the important groups and the threshold for a meaningful result before the experiment. Higher repeat purchasing may also reflect a competitor's shortage or a temporary price incentive. Context is necessary to interpret any number.

Look beyond the transaction. Did customers choose the suitable variant? Were expectations fulfilled? Did returns, complaints or service costs increase? What happened to margin and long-term experience? A brand that extracts more purchases through false urgency may seem more successful in the short run while damaging confidence in the longer run.

A serious study ends with more than 'our loyalty score improved'. It identifies which people, situations and decisions changed, what explanation is most plausible and which uncertainties remain. That makes associations a practical part of marketing strategy rather than another attractive metric with no operational consequence.

Customers may genuinely like a brand and still buy a competitor. That is not necessarily a broken relationship. Another use case, availability, variety, price or changed circumstances may explain the choice. Some positive associations remain intact while others become more important in a particular buying situation.

The answer is not to classify every brand switcher as a failed loyal customer. First examine which expectations the brand creates, when those expectations matter, whether it is an available option and what people actually do. Then make a change that addresses the diagnosed issue and test both behaviour and customer value.

A strong brand does not need to win every transaction. It needs to be credible, recognisable and available for the right need when the decision is made. That is the starting point for association-driven loyalty research.

Research limitation: These publications concern different sectors and methods, including retail, financial services and telecommunications. They do not establish why an individual in the fictional Zuidlicht case switches. The mechanisms described are hypotheses; the impact of a specific intervention requires appropriate testing.

• Associations in marketing: how brand meaning shapes choices
• The pack is still closed. Your brain has already judged the product.
• Your customer is satisfied. So why does the next order go elsewhere?
• Neurofactor - research methods