Most brand trackers report levels. The commercial signal sits in the gaps between them.

Most brand health reports move slowly and rarely disagree with themselves. Awareness is up two points. Favourability holds. Consideration is broadly flat. The report gets filed, the quarter continues, and nothing in it explains why revenue did something the brand data never signalled.

That is usually not a data problem. It is a reading problem. Almost everything on a standard tracker is a level, and levels are the least diagnostic output a tracker produces. A level tells you where the brand sits. It does not tell you where the brand is losing.

The teams that get commercial value out of brand measurement read the same reports differently. They treat each metric as one side of a ratio, and they spend their attention on the distance between adjacent stages rather than on the stages themselves. That distance is where demand leaks, and it is the part of the report that moves early enough to act on.

This piece sets out what brand health actually covers, which metrics are worth carrying, and how to tell the ones that govern a commercial outcome from the ones that merely reassure.

Brand health is the set of measures describing how a brand performs in the minds of the people who buy its category: awareness, consideration, preference, perception and loyalty, measured against competitors and tracked over time. The structural point most reporting misses is that these measures are only useful in relation to one another. A level tells you a position. A ratio between two levels tells you where demand is being lost.

Why a single brand health score tells you very little

The most common request we get from marketing teams is for one number. A brand health index, a composite score, something a board can watch move. It is an understandable ask and it produces a metric that cannot do the job.

A composite score averages measures that frequently move in opposite directions. Awareness climbs while consideration slips. Perception improves among existing customers while eroding among the ones you need next. Advertising awareness rises on the back of a campaign that shifted nothing further down. Average those together and you get a number that is remarkably stable, because the components cancel each other out.

Stability is precisely what makes it useless. A metric that cannot fall cannot warn you. Worse, it actively conceals the mechanism, because the moment you compress five signals into one you have thrown away the only information that told you which of the five was the problem.

The same failure shows up in single-metric reporting. A brand can hold high awareness for years while its commercial position collapses, and awareness will report the collapse as nothing at all. Fiat's Australian retreat is the cleanest recent example: near-universal recognition, negligible sales. Any tracker reporting awareness levels alone would have shown a healthy brand.

The gaps carry the diagnosis, not the levels

The useful unit of brand health is the conversion between one stage and the next.

Awareness to consideration. Of the people who know you, what share would actually consider buying you? This ratio isolates a relevance problem from a visibility problem, and the two have completely different remedies. A brand with 80 per cent awareness and 15 per cent consideration does not need more media. It needs a reason to be considered. A brand with 30 per cent awareness and 25 per cent consideration has the opposite situation and a much easier one to fix.

Consideration to preference. Of the people who would consider you, what share put you first? This is where competitive substitutability shows up. If consideration is healthy but preference is thin, you are on the shortlist and losing at the decision, which is a positioning and proposition problem rather than a funnel problem.

Preference to purchase. Of the people who prefer you, what share actually buy? A wide gap here usually points at something outside the brand: distribution, pricing, availability at the moment of choice. It is the ratio that most often exonerates the brand team and it is worth having on the record.

Enrolment to active use. For any membership or subscription structure, the gap between people signed up and people genuinely engaged is where margin quietly disappears. Loyalty programs are the standing example: membership charts rise whether the program is working or hollow, and only the ratio distinguishes the two.

Each of these ratios names a specific commercial failure and points at a specific fix. None of them is visible in a report of levels, which is why so many trackers produce interest without producing decisions.

What each brand health metric actually governs

The metric set below is close to standard. What matters is being explicit about the commercial question each one answers, because a metric carried without a question attached is a metric nobody acts on.

Unprompted awareness measures who comes to mind without help. It is the closest single proxy for what the Ehrenberg-Bass Institute calls mental availability, and the metric most sensitive to sustained investment. Treat first mention as distinct from total unprompted mention, because first mention tracks likely purchase choice far more closely than total recall does.

Prompted awareness measures recognition when the name is shown. It rises easily and it saturates. In most established categories it is a hygiene check rather than a growth metric, and reading it as a success measure is one of the more common ways brand reporting flatters itself.

Consideration measures the share of the market for whom you are a genuine option. Paired with awareness it produces the most useful ratio on the tracker. Ask about the competitive set at the same time, because knowing who else your considerers would consider is worth more than the consideration figure alone.

Preference measures how often you are the first choice among those choosing. It is the metric that most directly precedes share movement.

Brand attributes measure what the market believes you are. Their value is comparative, not absolute. An attribute score means nothing until it is set against competitors and against what the category says drives choice, which is a separate question and one worth answering with a forced trade-off rather than a rating scale.

Perception change asks whether views of the brand have improved or worsened, and over what. It is a leading indicator of churn risk and it is often the first metric to move after a service failure or a pricing decision.

Advocacy covers the willingness to recommend. Net Promoter Score is the common instrument and it is a reasonable cross-category benchmark, but it is a scoreboard rather than a diagnosis. The reasoning behind the score is where the value sits: why customers prefer you, and why the ones who left, left.

Advertising awareness attributes attention to channel and campaign. It is the metric that connects brand measurement to media decisions, and it is the one most often used to declare a campaign successful when nothing downstream moved.

Competitor comparison runs underneath all of the above. Every metric here is only interpretable relative to the set you compete with. A tracker measuring only your own brand produces numbers that cannot be judged.

How often should you measure brand health?

More frequently is not better, and the assumption that it is has sold a great deal of unnecessary tracking.

Brand metrics move slowly. Awareness and perception shift over quarters and years, not weeks. Measuring monthly against a sample that carries a meaningful margin of error mostly produces noise that teams then interpret as signal, which is worse than not measuring at all.

For most Australian organisations, a single annual wave with real strategic depth answers the questions that actually govern planning: where the brand sits, where it is losing, against whom, and what would need to change. Supplementary dips around a campaign or a market event are worth running when there is a specific decision waiting on the answer. That combination costs less than continuous tracking and produces more, because the depth goes into the wave where it earns something. We have set out what an annual report should actually answer separately, and the shape of that answer is what determines the cost of a tracking program far more than frequency does.

The exception is a category in genuine flux, or a brand mid-crisis, where the rate of change is itself the thing being measured.

What changes if you read the report this way

The practical test for any metric on your tracker is whether you can name the commercial decision it would change. If a number moved three points in either direction, what would you do differently? Metrics that survive that question are governing metrics. Metrics that do not are there to reassure, and reassurance is expensive when it is bought with research budget.

Most trackers carry too many of the second kind. They are easy to collect, they present well, and they rarely deliver bad news, which is exactly why they persist. The work of improving brand measurement is usually less about adding metrics and more about being ruthless with the ones already in the report.

Start with the ratios. Take the awareness and consideration figures you already have and calculate the conversion between them. Do the same for consideration and preference. If either is materially worse than your competitors' equivalent, you have found the constraint on growth, and it will be a more specific and more actionable finding than anything the levels told you. That single calculation is available to most marketing teams from data they already own, and it costs nothing.

Getting this right before 2027 planning locks matters more than usual, because the metrics chosen at the planning stage are the ones the year gets judged against.

Frequently asked questions

What is a good brand health score?

There is no absolute benchmark worth trusting, because scores depend on category structure, purchase frequency and how the questions were asked. A 40 per cent consideration figure is strong in a category with fifteen viable competitors and weak in one with three. The only meaningful comparisons are against your own competitive set, measured the same way, and against your own prior waves. Any provider quoting cross-category norms as a target is selling comparability that does not exist.

Can brand health be measured without a survey?

Partially, and less well than the alternatives suggest. Search volume, social listening and sales data all carry useful signal about behaviour. None of them tells you why a person did not consider you, what they believe about your brand, or who they chose instead. That gap is widening. IAB Australia reports that automated requests overtook human ones in web traffic during mid-2026, with most identified crawler activity now tied to AI systems rather than conventional search. As discovery moves inside answers that leave no behavioural trace, inference loses ground and asking people directly becomes more reliable, not less.

How large a sample do we need?

Enough to support the smallest subgroup you intend to make a decision about, which is the question that actually determines sample size. A national read is straightforward. A read by state, by age band, or among a low-incidence B2B audience requires materially more, because each cut needs its own defensible base. Design the sample from the decisions backwards, not from a headline total.

If your brand tracker is producing numbers that do not explain what your commercial results are doing, the problem is usually the metric set rather than the data. Brand Health designs custom research programs built around the ratios that govern growth in your category, rather than the levels that are easiest to report.

Schedule a free 30-minute consultation to discuss what your current brand metrics are not telling you.

Tom Morris is Managing Director of Brand Health, an Australian brand research consultancy specialising in custom brand tracking. Connect on LinkedIn.