What is brand image? And how do you research it?

You can describe your brand as knowledgeable, approachable and innovative. But does your audience see it that way? Existing customers might mainly associate it with reassurance. Potential customers might find it complicated. The difference between what you intend to communicate and what people connect with your brand is the starting point for brand image research.

A neuromarketing perspective considers more than the words people use. It also examines associations that are relatively easy to activate and the circumstances in which they become relevant. This calls for conversations, carefully chosen measures and a clear research question.

Brand image is the set of meanings, expectations and feelings that someone connects with a brand. Research also describes the patterns that occur within a defined audience. An association network provides a useful model for organising those connections. Keller, 1993 John et al., 2006

You investigate brand image by combining spontaneous associations and explicit ratings with suitable indirect measures, such as an Implicit Association Test. This allows different aspects of brand perception to be compared. No single method reveals the complete or uniquely 'real' image. De Houwer et al., 2009 Maison et al., 2004

Brand image describes the meaning a brand acquires for someone who encounters it. This can involve functional expectations, such as reliability, but also a feeling, a usage situation or the kinds of people thought to belong with the brand. Keller, 1993 John et al., 2006

For a fictional software brand, possible associations might include 'overview', 'control' and 'complicated'. These are illustrations, not research findings. The words alone do not explain why someone mentions them. Control could appeal to a person responsible for finance while feeling restrictive to an employee. Investigate meaning and context before coding an association as positive or negative.

A brand does not necessarily have the same image for everyone. Describe patterns within a particular group rather than relying on an average that conceals differences. In B2B, consider users, technical advisers, procurement professionals and budget holders. Research in business markets shows that buyers differ in the importance they attach to brands. Mudambi, 2002

Brand image, brand identity and brand awareness

This article uses the following distinctions:

Concept	Central question	What to keep separate
Brand identity	Who are we and what do we want to stand for?	Internal principles, behaviour and intended presentation.
Brand image	What does the brand mean to the audience?	Perceived and felt meaning, rather than intention alone.
Brand awareness	Is the brand recognised or recalled?	Being known does not establish favourable associations.
Brand perception	How does someone perceive and interpret the brand?	An overlapping term often used interchangeably with brand image.

The distinction between awareness and image follows Keller's model of brand knowledge. The other formulations are working definitions for this article. We do not impose a strict rule that perception is always individual and image always collective. Keller, 1993

Imagine a brand name as a starting point for connections with other concepts. Thinking of a museum might connect with painting, a distinctive building or an afternoon out. An association network helps represent these relationships. It is a cognitive model and a research representation, not a picture of directly measured connections between brain cells. Keller, 1993 John et al., 2006

The question is not only which association is present, but how it connects with other meanings. 'Exclusive' might connect with quality for one group and with distance or inaccessibility for another. Investigate those relationships before placing similar words in the same cluster. See Association: meaning and application and What are association clusters?.

In your research plan, distinguish an association's content, its positive or negative evaluation, its accessibility in a particular situation and the extent to which people share the pattern. This is a practical organisation of research questions, not a universally validated overall brand image score. When discussing association salience, state what you measure: for example, spontaneous availability or performance in a specifically designed task. A frequently mentioned word is not automatically important to a purchase decision.

Explicit data are directly expressed answers, such as a reliability rating. Implicit measures infer an evaluation or association indirectly from task performance. The distinction concerns measurement. 'Implicit', 'automatic' and 'unconscious' are not interchangeable terms. De Houwer et al., 2009 Greenwald et al., 2022

The psychological foundations of brand image

Repetition and familiarity. Repeated exposure can affect recognition, familiarity and liking. More exposure does not produce unlimited increases in liking. Repetition is therefore no substitute for a relevant message or a good experience. Montoya et al., 2017

Emotional pairings. Evaluation can change when something is paired with positive or negative stimuli. This is called evaluative conditioning. The existence of such effects does not establish that the learning process necessarily takes place entirely outside awareness. Hofmann et al., 2010

Social meaning. People also form brand meanings with others; see also social proof. Research on brand communities describes how social relationships, shared stories and practices become part of a brand's meaning. Muñiz & O'Guinn, 2001

Personal experience. Brand experience extends beyond advertising. Encounters with products, services and their usage environments matter too. Research links brand experience with satisfaction and loyalty, without making experience and image the same concept. Brakus et al., 2009

The practical question is therefore: where does an association come from? Investigate whether someone is describing their own experience, repeating something they heard or responding mainly to the current presentation. Its origin helps identify where an improvement should be made.

A favourable image can provide reasons to consider a brand. Research on brand credibility identifies relationships with the brands people consider and choose. This supports the relevance of trust, but does not demonstrate that every positive association causes a purchase. Erdem & Swait, 2004

Price perception also deserves its own research question. A credible brand may be assessed differently when prices vary; research on price sensitivity finds differences between product categories. That does not establish a fixed percentage that your brand can add to its price. Erdem et al., 2002

Loyalty likewise requires several explanations to be distinguished. Brand experience is related to satisfaction and loyalty, but repeat purchasing could also involve convenience, availability or contract terms. Alongside evaluation, investigate reasons for staying and actual choice behaviour. Brakus et al., 2009

For positioning, a useful question is: which meaning can you credibly establish, for whom and in which situation? Emotion need not be treated as the opposite of business reasoning. Take a possible B2B association such as 'reassurance' as a starting point to investigate whether it concerns continuity, reduced personal risk or predictable support. This is a research hypothesis, not an established pattern for all business buyers. Mudambi, 2002

Start with the decision you need to inform, not the technique. Do you want to understand what your brand stands for, how it differs from a competitor or whether new positioning changes anything? These questions do not automatically require the same measure.

The overview below is a practical synthesis of association research, self-report and implicit measurement approaches. John et al., 2006 Nisbett & Wilson, 1977 Greenwald et al., 1998 De Houwer et al., 2009 Maison et al., 2004

Method	Useful research question	What the result does not establish on its own
Open associations, interviews and focus groups	Which meanings, experiences and words emerge?	That a stated reason fully explains the underlying choice process.
Surveys using explicit ratings	How does the audience evaluate selected characteristics?	That those characteristics are spontaneously available or accessible during purchase.
An Implicit Association Test and other suitable response-time tasks	Which pairings are relatively easier in the selected task?	An absolute, context-free or complete unconscious brand image.
Choice measures and observation of behaviour	Are the insights related to what people choose or do?	That brand image is the cause without a suitable research design.

A workable research design

1. Define the audience and context. Specify whom you are studying, what brand experience is required and which decision matters. Distinguish, for example, existing customers from people who only recognise the name. In B2B, plan which decision-making roles require separate attention.

2. Explore without supplying the answers. First ask openly what the brand brings to mind and investigate concrete experiences. Only then introduce a list of desired characteristics. This at least avoids placing your own brand vocabulary in the opening question.

3. Turn the main questions into measurable comparisons. Select the associations to investigate and the tasks to use. Define comparison brands, instructions, outcome measures and exclusion rules in advance. Check whether words and images are understandable and appropriate. Determine sample size from the required precision, comparison and analysis; do not assume one standard number fits every study. Greenwald et al., 2022

4. Combine findings without treating them as equivalent. Compare spontaneous answers, direct ratings, indirect task results and choice data. Report variation and uncertainty. A discrepancy can prompt questions about context, task design or different meanings. It does not establish that a participant is lying.

5. Test a specific change. Translate an insight into one clearly described improvement and measure again under comparable conditions. A change after a campaign does not establish that the campaign caused it. Plan a suitable comparison group or a randomised comparison when a causal conclusion is needed.

This process is editorial research guidance, not a separately validated Neurofactor instrument. See also How do you measure associations?.

Traditional methods: the value and limits of conscious reports

Surveys, interviews and focus groups are important for understanding language, experiences and arguments. They can help prevent researchers from imposing their own interpretation on a measurement.

However, an answer to 'why did you choose this brand?' is not a complete recording of the mental choice process. People may offer a plausible explanation without knowing every influence on their decision. Nisbett & Wilson, 1977 Social desirability can also matter for sensitive questions. Its relevance varies by question and situation; this is not a reason to distrust every brand rating. Tourangeau & Yan, 2007

Make questions concrete. Ask about a recent experience, allow 'I don't know' and collect individual responses before a group discussion. Investigate free associations before recognition of suggested characteristics. These are practical design choices, not guarantees that every source of bias disappears.

Implicit measures within neuromarketing

Implicit association research adds a different type of observation. In a classic IAT, participants categorise words or images as quickly as possible. Differences between combinations provide information about relative associations within that task. Greenwald et al., 1998

An illustrative design could compare two brands with concepts related to 'reliable' and 'unreliable'. The question is whether one pairing is processed relatively more easily than the reversed pairing. This is not equivalent to determining someone's absolute trust in a single brand. The comparison categories are part of the interpretation. Greenwald et al., 1998 Greenwald et al., 2022

Across three consumer studies, Maison and colleagues found that IAT measures added information beyond explicit ratings when investigating brand attitudes and behaviour. This demonstrates potential added value, not universal predictive power for every customer or product. Maison et al., 2004

Response time is a measurement, not a ready-made interpretation. Task structure, stimulus familiarity, errors and scoring require consideration. A fast response alone does not show whether an association is positive. Nor does an indirectly measured association establish that the participant cannot be aware of it. De Houwer et al., 2009 Greenwald et al., 2022

Although used within neuromarketing, an IAT is a behavioural, psychological measure rather than a direct brain measurement. Additional brain or eye-movement measurements can address different questions about processing a brand communication. They do not directly read a complete brand image or a specific emotion. Choose them only when they add something to the question being investigated. Plassmann et al., 2012

Practical example: the Mauritshuis and the meaning of the experience

On 2 October 2024, the Mauritshuis published research (see the Mauritshuis case) on experiencing art. The museum commissioned Neurofactor to conduct the research. Among other things, it examined experiencing original artworks in the museum environment and reproductions. Mauritshuis, 2024a Mauritshuis, 2024b

The example is relevant to brand image because the museum used it to communicate the value of encountering original art. A possible implication for brand strategy is to connect an organisation's promise with the experience it actually offers. That implication is an interpretation here, not a measured image effect. Mauritshuis, 2024a

The published research is not a direct brand image study. It does not establish that the association network surrounding the Mauritshuis changed, that the brand became more attractive or that the research increased visitor numbers. The public document is a project report, not a peer-reviewed journal article. Mauritshuis, 2024b

A suitable follow-up could investigate the associations visitors and non-visitors connect with the Mauritshuis, how they evaluate the promise of original art and whether communication about it changes relevant associations. This is a proposal for further research, not a description of an already completed project.

The following recommendations translate research questions into brand choices. Treat the expected effects as hypotheses to test.

Choose a meaning you can deliver. Do not replace broad terms such as 'quality' with another vague promise. Formulate a clear core message. Specify the experience you want the audience to connect with it and the evidence you can provide.

Align communication with experience. Investigate whether the website, sales conversation, service and product usage support the same expectation. Address a recurring service problem before merely creating a new campaign.

Use emotional stories with a recognisable brand connection. Ask not only whether a story moves people, but what they subsequently connect with the brand. An appreciated story and a more favourable brand image are different outcomes.

Protect what already works. Before repositioning, record the associations you intend to retain, strengthen or change. A new design should support the intended meaning, not simply be surprising.

Turn the pattern into testable choices. Within Neurofactor's approach, an associative signature can help describe a recurring pattern. Do not treat this working concept as automatic evidence of behaviour. Connect the pattern with specific communication strategy choices and test them. See How do you turn a target group profile into a communication strategy?.

Presenting only your desired characteristics. This investigates recognition of your vocabulary while unexpected associations may remain hidden. Explore openly first.

Treating either explicit or implicit data as the truth. Value comes from suitable methods and their comparison, not the assumption that one technique replaces every other. De Houwer et al., 2009 Maison et al., 2004

Confusing frequency, evaluation and importance. 'Expensive' could describe a price impression, an objection or a sign of exclusivity. Preserve original responses and investigate meaning before combining them.

Averaging everything and concealing differences. Separate relevant groups and show uncertainty. Do not immediately label an incidental subgroup a new market segment.

Presenting a snapshot as a permanent image. Record the measurement date, audience, context and recent encounters. A repeat measurement can only be interpreted when it is clear what remained consistent and what changed.

Jumping from association to causation. Greater brand preference after an intervention can be an important signal. The intervention's contribution must be supported by the research design, not the persuasiveness of the presentation.

Your brand image is not simply what you say about yourself, but what people connect with your brand. Good research clarifies which meanings exist, whom they concern and which questions remain open.

The value of neuromarketing here lies in combining perspectives: what people can articulate, what suitable indirect tasks show and what happens in relevant choices. This turns a general judgement about your brand into a focused research question. Rather than 'is our image good?', ask: 'which associations help or hinder this audience in this situation, and which improvement can we substantiate?'