Brand equity is one of the most cited concepts in marketing and one of the least understood in practice. Most brand teams can tell you their NPS score, their aided awareness number, and their latest Net Sentiment from social listening. Very few can tell you what drives those numbers, which ones actually predict growth, or what they would do differently if those numbers changed.

That's not a data problem. It's a research design problem.

This article breaks down what brand research is, which methods and models are actually worth using, and how to build a research program that produces insights your team can act on.

What brand research is actually measuring

At its core, brand research tries to answer one question: what does your brand mean to people?

Not whether they've heard of you. Not whether they'd recommend you to a friend on a scale of 1 to 10. What your brand actually means — the associations, feelings, and memories that surface when someone encounters it. Whether it comes to mind in situations that lead to a purchase. Whether it feels different from the alternatives.

David Aaker, whose work on brand equity is still the most useful framework in practice, described brand value as the premium a brand generates beyond what the bare product would command. That premium is built from five assets: awareness, associations, perceived quality, loyalty, and proprietary assets like patents or symbols. Each is measurable. Each can erode without a team noticing until the financial effects show up quarters later.

The reason most brand research fails is that it measures outputs (scores) rather than the drivers behind them. A declining awareness metric tells you something changed. It doesn't tell you what changed or what to do about it. Getting to the why requires a different kind of research than most teams are running.

The methods that actually tell you something

Brand research methods split into qualitative and quantitative, but the more useful distinction is between methods that surface new information and methods that validate whether what you already know holds at scale.

Qualitative methods surface. In-depth interviews, conducted properly, reveal the emotional logic behind brand choices — why someone feels comfortable with one bank and vaguely uneasy about another, why a skincare brand feels like it's "for them" while a functionally identical competitor feels like it isn't. Focus groups can be useful for watching how people react to a stimulus in real time, but they're unreliable for opinion measurement. Group dynamics do strange things to stated preferences. The loudest voice in the room shapes the others, and what people say in a group often diverges from what they actually think alone.

Diary studies are underused and underrated. Asking participants to self-document their experiences with a product or category over days or weeks surfaces the texture of everyday behavior that a one-time interview misses. The frustration someone forgot by the time they came in for the session. The context in which they actually use the product, which turns out to be nothing like the usage occasion the brand assumed.

Quantitative methods validate. Surveys measure how widespread something is, track whether it's changing over time, and allow for statistical comparison across segments. Brand tracking — a recurring survey program measuring awareness, consideration, perception, and preference — is the operational heartbeat of a brand health program. The problem most teams run into is running tracking surveys without any qualitative layer to explain the numbers. When perception of a brand attribute drops three points, the tracker surfaces the symptom. Understanding the cause requires talking to people.

MaxDiff and conjoint analysis are worth knowing. MaxDiff asks respondents to rank attributes against each other rather than rate them on a scale, which eliminates the rating bias that makes most attribute batteries useless. Conjoint forces respondents to make trade-offs that resemble real purchase decisions. Both are better tools for measuring what customers actually value than asking them to agree or disagree with a list of statements.

The four frameworks worth knowing

Brand researchers argue about models the way product managers argue about frameworks. The reality is that the model matters less than consistency. Pick one that fits your business, use it every time, and you'll accumulate data that tells you something over time.

That said, the four models that show up most often are worth understanding.

Aaker's Five-Asset Model is the most practical for internal use. It gives you a clear checklist of what to measure and a language for talking about brand health across functions. If your brand team and your finance team are ever going to be in the same conversation about brand investment, Aaker gives you the vocabulary.

Keller's CBBE pyramid is the most intuitive for explaining brand building to people who aren't researchers. It starts with awareness and moves through association, quality, and response to resonance — the state where customers feel a deep, almost irrational attachment to a brand. Most brands are not at resonance. The model helps you identify where you actually are.

Young & Rubicam's Brand Asset Valuator is less widely used but contains a genuinely useful insight: that differentiation and relevance are not the same thing, and that brands often decline in one before the other. A brand can remain highly relevant to existing customers while losing differentiation against new entrants. The BAV grid catches that earlier than most other models.

Kantar's BrandZ is tied most closely to financial valuation and is designed for cross-category, global benchmarking. It's not the right tool for diagnosing a specific brand problem, but it's useful if your question is about relative brand value against competitors across markets.

What most brand trackers are missing

The most common failure mode in brand research is running a tracker for years and not knowing why anything is changing.

A well-designed tracker measures the right metrics at the right frequency. But the number itself is never the insight. "Perceived quality up 2 points" is a data point. Understanding which customers changed their minds, what drove the change, and whether it reflects a real shift in experience or just a response to recent advertising — that's the insight. Getting there requires qualitative work alongside the quantitative tracking.

The brands that get the most from their research programs treat tracking as a signal for where to dig, not the end of the process. When a metric moves, they run interviews. When awareness drops in a segment, they go talk to that segment. When a competitor starts closing the differentiation gap, they find out why.

Corvane is built for exactly this kind of iterative program. Because research runs on your own internal panel — opted-in customers who have real context on your brand — you can move from a quantitative signal to a qualitative follow-up study in days rather than weeks. The Co-pilot synthesizes patterns across hundreds of open-ended responses automatically, so you're not waiting for an analyst to manually code 300 survey verbatims to find out what's actually driving a number.

The metrics that matter

Unaided awareness in context. Asking "have you heard of Brand X" measures recognition memory. Asking "when you think about buying a moisturizer, which brands come to mind" measures recall in a purchase-relevant situation. The second is far more predictive of actual consideration and should be the primary awareness metric for most brands.

Mental availability across category entry points. This is Byron Sharp's contribution to the field, and it's more useful than traditional awareness tracking. Instead of measuring whether people know your brand in the abstract, you measure whether your brand comes to mind across the range of situations, occasions, and triggers that actually drive category purchases. A brand with strong mental availability is present in more of those moments.

Brand differentiation. Not just "is our brand different" but "do consumers see it as meaningfully different in ways that affect their choice." Distinctiveness (being visually or stylistically recognizable) is not the same as differentiation (being chosen for a reason). You need both. Distinctiveness gets attention. Differentiation closes the sale.

Brand associations. The network of attributes, feelings, occasions, and meanings that consumers connect to your brand. These are built slowly and erode faster than most teams expect. Strong associations are specific — not "premium" in the abstract but "the brand that costs more and is worth it" in a specific category context. Weak associations are generic, interchangeable, and don't drive choice.

Trust, where it matters. In categories where the purchase decision carries risk — healthcare, financial services, food — trust is a primary driver of brand preference. In categories where it isn't, trust functions more as a hygiene factor: its absence hurts you, but its presence doesn't differentiate you. Know which category you're in before investing heavily in trust metrics.

How to run a brand research project that actually produces something useful

Start with a decision, not a question. The difference sounds small but changes everything. "How is our brand perceived?" is a question that produces a report. "We're considering entering the premium segment and need to know whether our current associations help or hurt that move" is a decision that produces a recommendation. Every research brief should name the decision it's designed to support.

Be honest about your sample. Brand research that only includes your existing customers will produce a flattering but incomplete picture. Your loyal customers think well of you. That's why they're still customers. What you need to understand is why non-customers haven't chosen you, what lapsed customers found wanting, and what category buyers associate with your brand compared to competitors. Those are different samples with different recruiting requirements.

Segment before you analyze. Different customer groups hold genuinely different perceptions of the same brand simultaneously. A brand can be seen as trustworthy by its core demographic and as stuffy or inaccessible by a younger segment it's trying to reach. Collapsing those into a single average produces a number that describes no one. Corvane's audience segmentation tools let you define your segments before the survey goes out, so the analysis reflects the actual structure of your customer base.

Connect the metric to the driver. A three-point drop in perceived differentiation is a problem. A three-point drop in perceived differentiation driven primarily by customers in the 25-34 segment who have recently tried a specific competitor is a different problem with a different solution. Get to the driver.

Common questions

What's the difference between brand research and market research? Market research is about the landscape: how big is the category, who are the competitors, what does the demand curve look like, where is the growth. Brand research is about your place in that landscape: what do people think of you, why do they choose you or not, what would it take to change their minds. Both are necessary and neither substitutes for the other.

What's the difference between brand tracking and brand research? Tracking is one tool within brand research. It measures health metrics at regular intervals. Brand research is the broader practice of understanding why your brand is perceived the way it is. Running a tracker without any deeper research is like checking your temperature every day without ever seeing a doctor. You know something's wrong. You don't know why.

How often should you run brand research? Foundational studies annually, with tracking quarterly. That cadence is the minimum for catching meaningful shifts before they compound. The teams that get the most value from research run it more frequently, because each study informs the next and because decisions get made faster when the data is recent enough to be trusted.

What's the difference between brand loyalty and brand preference? Preference is stated: given a choice, I'd pick this brand. Loyalty is behavioral: I keep picking this brand, including when a cheaper or more convenient option is available. Most brands overestimate their loyalty by measuring preference. True loyalty is rarer than brand teams usually admit, and measuring it requires behavioral data, not just survey responses.

Will AI replace brand researchers? The honest answer is: the parts of brand research that are slow, expensive, and repetitive are already being automated. Transcription, theme clustering, pattern detection across large response sets — these are tasks AI handles faster and more consistently than humans. What it can't do is design the right study, recognize when an unexpected finding matters, or make the judgment call about what a result means for the brand's strategy. The researchers we see getting the most out of Corvane are the ones who use the AI synthesis to do in minutes what used to take days, and spend the time they've saved on the judgment work that actually requires a human.