Branding Research: What It Should Tell You and What It Usually Doesn’t

Branding research is the process of gathering structured insight about how a brand is perceived, by whom, and against what alternatives. Done well, it tells you where you stand, where the gaps are, and what a credible positioning could look like. Done poorly, it produces a thick deck of charts that confirms what the client already believed and gets shelved within a month.

Most branding research falls somewhere in the middle: methodologically defensible, commercially useless. The problem is rarely the data. It’s the questions being asked before the research begins, and the critical thinking applied to the findings after it ends.

Why Most Branding Research Fails Before the Fieldwork Starts

I’ve sat in enough research debrief sessions to know the pattern. The agency presents the findings. The room nods. Someone says “that’s really interesting.” And then nothing changes, because the research answered questions nobody was actually arguing about.

The failure usually happens at the briefing stage. Branding research gets commissioned when a business is about to reposition, launch into a new market, or refresh its identity. Those are legitimate triggers. But the brief often says something vague like “we want to understand how we’re perceived in the market.” That’s not a brief. That’s a statement of curiosity. And curiosity doesn’t have a budget attached to it.

Good branding research starts with a specific commercial question. Not “how are we perceived?” but “why are we losing mid-market clients to Competitor X despite having a stronger product?” or “what do procurement teams actually value when they’re shortlisting agencies, and how does that match what we’re saying about ourselves?” Those questions have answers that change decisions. The vague ones don’t.

If you’re thinking about where branding research sits within a broader strategic process, the brand positioning and strategy hub covers the full architecture, from audience work to competitive mapping to positioning frameworks. Research is one input into that process, not the whole thing.

What Types of Branding Research Actually Exist?

There are four broad categories of branding research, and most projects should draw on at least two of them. Using only one is like handling with half a map.

Brand Awareness and Salience Research

This measures whether people know your brand exists, and whether it comes to mind in the relevant buying context. There’s a meaningful difference between aided awareness (do you recognise this brand?) and unaided awareness (which brands come to mind when you think of X?). Unaided awareness is the harder number to move and the more commercially relevant one.

Salience, which is the probability of being thought of in a buying situation, is arguably more important than awareness on its own. A brand can be widely known and still not be considered. Wistia makes this point well: awareness metrics can look healthy while conversion and consideration numbers are quietly eroding. That gap is worth investigating.

Brand Perception and Association Research

This is the most commonly commissioned type and the most commonly misread. It tells you what attributes people associate with your brand, how those compare to competitors, and where the perception gaps are relative to your intended positioning.

The misreading usually happens when teams treat association scores as fixed truths rather than snapshots. Perceptions shift with product experience, media exposure, word of mouth, and broader category trends. A brand that scores well on “trustworthy” in a stable market can see that score deteriorate quickly when a single high-profile failure occurs, as BCG’s work on brand advocacy illustrates. The number at a point in time is useful. The trajectory over time is more useful.

Brand Loyalty and Retention Research

This goes beyond perception into behaviour. Are people buying again? Are they recommending? What would make them switch? Loyalty research is particularly valuable in categories where switching costs are low, because the gap between stated loyalty and actual behaviour tends to be wide. MarketingProfs has documented how brand loyalty weakens under economic pressure, which tells you something important: loyalty scores gathered in stable conditions may not predict behaviour when conditions change.

Qualitative Brand Research

Focus groups, depth interviews, and ethnographic observation. These don’t give you statistical confidence, but they give you something quantitative research rarely does: the language people use, the mental models they apply, the emotional texture of how they relate to a category. That texture is what makes the difference between a positioning that sounds right in a boardroom and one that actually resonates with a real person.

I’ve always found that qualitative research is most useful before you design the quantitative instrument, not after. If you run the survey first, you’re testing hypotheses you invented. If you run the qual first, you’re testing hypotheses the market gave you. That’s a meaningfully different starting point.

How Do You Read Branding Research Without Being Misled By It?

I don’t reject research. I don’t think it’s theatre or a waste of budget. But I don’t accept it uncritically either, and I’ve seen too many senior teams treat a research deck as settled truth when it was, at best, an informed estimate.

The questions I apply to any piece of branding research are fairly consistent.

First: who was sampled, and does that sample actually represent the people who matter to this business? A B2B brand that surveys consumers, or a premium brand that surveys a general population panel, is measuring something real but not something relevant. Sample design is where a lot of research goes quietly wrong without anyone flagging it.

Second: are the differences statistically meaningful or just numerically different? A brand scoring 62% on “innovative” versus a competitor at 58% is not a strategic insight. It’s noise. I’ve watched teams build entire positioning pivots on differences that sat comfortably within the margin of error. The number looked decisive on a slide. It wasn’t.

Third: what are people actually reporting, and does that match their behaviour? Stated preference and revealed preference diverge constantly in brand research. People say they value sustainability, then buy on price. They say they prefer local brands, then choose the global one because it’s easier. The gap between what people say and what they do is not a research failure. It’s a human reality. Good research design accounts for it. Bad research design ignores it and reports the stated preference as if it were reliable.

Fourth: is this insight or just drama? A finding that “only 34% of respondents could correctly identify our brand from a logo” sounds alarming. But if your brand is three years old, operates in a niche B2B category, and has never run above-the-line advertising, that number is completely expected. Context is everything. Research without context produces anxiety, not strategy.

What Does Branding Research Actually Inform?

This is where I think a lot of teams get confused. They commission branding research and then use it to justify decisions that were already made, or apply it to questions it wasn’t designed to answer.

Branding research should inform four things specifically.

Positioning decisions. If you’re choosing between two credible positioning territories, research can tell you which one has stronger existing associations to build on, which one faces more competitive crowding, and which one resonates more clearly with the audience segments that matter most commercially. That’s a legitimate use of research data.

Messaging hierarchy. Knowing which attributes your audience values most, and which ones they currently associate with you versus competitors, helps you decide what to lead with. This is particularly useful when you’re trying to shift perception in a specific direction rather than simply reinforce what already exists.

Visual and tonal identity decisions. Research can validate whether your current identity is being read the way you intend, and whether a proposed direction feels coherent with your positioning. Visual coherence matters more than most marketers acknowledge, and research can surface disconnects between what the brand says and what the identity signals.

Competitive strategy. Understanding how competitors are perceived, where they’re strong, and where they’re vulnerable gives you a map of the positioning space. You can choose to compete directly on the same territory, or find a credible space that’s less contested. BCG’s thinking on brand and go-to-market alignment is relevant here: brand positioning doesn’t exist in isolation from commercial strategy, and research that maps both together is more useful than research that treats brand perception as a standalone variable.

What branding research should not be used for: making creative decisions by committee, validating a rebrand that’s already been signed off, or producing a number that makes the marketing team feel like they’ve done their due diligence. I’ve seen all three. None of them produce good outcomes.

The Specific Research Methods Worth Knowing

Beyond the broad categories, there are specific methodologies that come up repeatedly in branding work. Each has a legitimate use case and a common misuse.

Brand tracking studies measure perception and awareness over time, usually quarterly or annually. Their value is in the trend line, not the snapshot. A single wave of tracking data tells you where you are. Three or four waves tell you whether you’re moving, and in which direction. Many businesses commission one wave, declare the research “done,” and never track again. That’s like taking a single blood pressure reading and concluding you understand your cardiovascular health.

Conjoint analysis is used to understand how people trade off different brand attributes when making choices. Instead of asking “how important is quality to you?” (everyone says very important), it presents realistic choices that force trade-offs. The output is a much more honest picture of what actually drives decisions. It’s more expensive and more complex to run, but in categories where purchase decisions involve multiple competing factors, it’s worth the investment.

Net Promoter Score is widely used and widely misunderstood. It measures the likelihood of recommendation, which is a useful proxy for brand loyalty and satisfaction. What it doesn’t tell you is why people are or aren’t likely to recommend, or what would change their score. Used as a standalone metric without diagnostic follow-up, it produces a number that looks like insight but isn’t. Moz’s analysis of local brand loyalty points to the importance of understanding the specific drivers behind loyalty scores rather than treating the score itself as the answer.

Semiotics and cultural analysis sits at the edge of what most marketers would call “research,” but it’s genuinely useful for understanding the broader cultural codes that shape how a brand is read. What does your category signal culturally? What visual and linguistic conventions do audiences expect, and which ones are available for differentiation? This kind of analysis tends to be qualitative and interpretive, which makes some clients uncomfortable. But it surfaces things that surveys simply can’t.

Digital and behavioural data is increasingly being used as a complement to traditional brand research. Search behaviour, social listening, review sentiment, and content engagement patterns all tell you something about how a brand is being perceived and discussed in the real world, without the filter of a survey question. The limitation is that it captures what people do publicly, which skews toward the vocal minority. It’s a useful signal, not a representative sample.

When Should You Commission Branding Research?

There are four moments when branding research genuinely earns its budget.

Before a major repositioning. If you’re about to make a significant strategic shift, you need a baseline. Where are you now? How are you perceived? What associations do you own? Without that baseline, you can’t measure whether the repositioning worked, and you’re more likely to make assumptions about your starting point that turn out to be wrong.

Before entering a new market or segment. What you know about your brand’s perception in your existing market may not translate. Different audiences, different competitive contexts, different cultural codes. Research before entry is significantly cheaper than a failed launch followed by research to understand why it went wrong.

When performance data shows something unexpected. If conversion rates are dropping, if win rates in sales are declining, if customer retention is softening, brand perception research can help diagnose whether the problem is a brand issue or something else entirely. I’ve seen businesses spend months optimising their sales process when the real problem was that their brand had developed a reputation for poor delivery. The sales team couldn’t fix that. The brand team could.

After a significant brand event. A major product failure, a public controversy, a merger or acquisition, a high-profile leadership change. These events shift perception, sometimes dramatically and sometimes in ways that aren’t visible in commercial data until the damage is done. Moz’s piece on risks to brand equity touches on how quickly brand equity can erode when external events interact with existing perception vulnerabilities. Research after a significant event tells you what you’re actually dealing with, rather than what you assume.

The moment when branding research is least useful is when it’s commissioned as a substitute for a strategic decision. If the leadership team can’t agree on the direction, research rarely resolves that. It produces more data for both sides to interpret in their preferred direction. The disagreement is usually about values and priorities, not about facts. No amount of fieldwork fixes that.

The Gap Between Research and Strategy

When I was running an agency and we were working on positioning for a client in a competitive B2B category, we commissioned a fairly standard brand perception study. The findings were clear: our client was seen as reliable but not innovative. Their competitors owned the “forward-thinking” space. The obvious strategic response was to try to shift perception toward innovation.

But when we went back to the client’s commercial data, we found something the research hadn’t surfaced. Their highest-value clients, the ones with the longest tenure and highest spend, specifically chose them because of their reliability. They weren’t looking for innovation. They were operating in a risk-averse environment and they valued a supplier who did what they said they would do, consistently, without drama. The “innovation gap” was a gap relative to competitors. It wasn’t a gap relative to what their best clients actually wanted.

The research was accurate. The strategic conclusion most people would have drawn from it was wrong. That’s the gap I’m talking about. Research tells you what is. Strategy decides what matters.

Wistia’s analysis of why brand building strategies fail makes a related point: the problem is often not the brand itself but the assumptions baked into the strategy that surrounds it. Research can surface the symptoms. It takes commercial judgment to diagnose the cause.

If you want to see how branding research fits into a complete strategic process, from initial diagnosis through to positioning, architecture, and activation, the brand strategy section of The Marketing Juice covers each stage in detail. Research is one of the earlier inputs. What you do with it is the harder part.

About the Author

Keith Lacy is a marketing strategist and former agency CEO with 20+ years of experience across agency leadership, performance marketing, and commercial strategy. He writes The Marketing Juice to cut through the noise and share what works.

Frequently Asked Questions