Brand Health and Equity Analysis

Why does my brand score lower than the market leader on every image attribute?

Most likely because they are bigger, not because they are better regarded. Larger brands score higher on almost every attribute, including ones they have no logical claim to, because more people know them and familiarity generates agreement. Before concluding anything about perception, recompute the attributes among those aware of each brand, index each attribute against the brand's own average, and where you have several brands, regress attribute score on awareness or penetration and read the residual. The residual is the honest measure of distinctive association.

What is the most useful number in a brand funnel?

The conversion between adjacent stages, not the level at any one. Two brands with identical consideration can have completely different problems: one converts a large aware base poorly, which is a meaning problem, and the other converts a small aware base extremely well, which is a reach problem. Those two diagnoses lead to opposite investments, and only the conversion ratio distinguishes them.

Why do funnel bases matter so much?

Because each stage sits on a different denominator, and the denominators differ by brand. A brand with 90% awareness has a far larger consideration base than one with 40%, so a side-by-side funnel chart across brands is comparing figures whose bases are not comparable. Compounding it, conversion ratios between two small numbers are the least stable statistic in the format: a considering base of 45 will not support a reliable considering-to-preferring ratio, however clean the arithmetic looks.

How should I read a perceptual map?

Carefully, and with four things stated. The axes are statistical dimensions, so naming them is human interpretation and should be shown as such. Brand-to-attribute proximity is read directionally from the origin, not as a simple distance score, and brand-to-brand distance does not mean the same thing. The map shows only the variance the two plotted dimensions capture, so report that proportion. And a brand near the origin is not average in a good way: it usually means undifferentiated, though it can also mean too few people rated it.

What is the difference between awareness and salience?

Awareness is recognition: whether someone knows the name. Salience, in the mental availability framing, is retrieval: whether the brand comes to mind in the buying situations that matter. A brand can have very high prompted awareness and be retrieved for almost no occasions, which is a specific and fixable problem that aggregate awareness hides entirely. Prompted agreement that a brand suits an occasion is a much weaker signal than unprompted retrieval, and the two should never be presented as the same measure.

Do brand health metrics predict sales?

That is an empirical question about your category and your brand, and it usually has not been tested. Where you have several waves and commercial data, correlate the series and report what you find, along with the common causes that were not controlled: distribution, price, seasonality, category growth, competitor activity. A few waves is not enough to estimate a relationship. Where the link has not been established, every claim about business consequence built on a brand metric is an assumption, and it should be labelled as one.

Which brand attributes are worth investing in?

Those with a large spread across brands in the category, where ownership is possible. Attributes on which every brand scores similarly are hygiene: a low score is a problem, but a high score is not an advantage, and building an association everyone already has returns nothing. Compute the spread per attribute before any recommendation, then look at your brand's size-adjusted residual on the high-spread attributes.

Can I compare brand results across markets?

Not on raw levels. Attribute agreement, scale use and advocacy measures vary with cultural response style at least as much as with brand perception, so a lower score in one market may be a response-style artefact. Compare within-market standardised scores or within-market ranks, and treat the interpretation as a judgement requiring local knowledge rather than an analytical output.

Where do brand category norms come from?

Often from nowhere checkable, which is a problem given how much weight they carry in a brand debate. Before quoting a benchmark, know its source, its category, the exact question wording behind it and its date. A norm from a different category, a different question or a different decade is not a benchmark, and an unsourced one should be treated as a fabricated source rather than a soft fact.

Can AI analyse brand tracking data reliably?

It handles the base arithmetic and the size correction consistently, which is genuinely valuable given how often both are done wrong by hand. The characteristic risks are importing category norms with no source, naming perceptual map axes confidently as though the labels were results, building conversion ratios through stages that were not routed, and letting the size-confounding caveat drift off the image chart into an appendix. Require the metric definition table first, the routing inspected in the data rather than assumed, and every caveat on the same page as the number it qualifies.