Brand equity

Origin

The term arrived through the finance department. Keller (1993) opens with two motivations for studying brand equity: a financial one, “to estimate the value of a brand more precisely for accounting purposes” or for mergers and acquisitions, and a strategic one, to improve marketing productivity [2, p. 1]. Acquirers in the 1980s were paying multiples of book value for consumer companies, and nobody could say what the premium bought.

The academic conversation was formalised at a Marketing Science Institute conference in Austin, Texas, on 1 to 3 March 1988, summarised as MSI Report 88-104, “Defining, Measuring and Managing Brand Equity” [4]. Farquhar (1989) supplied the working definition: brand equity is “the ‘added value’ with which a brand endows a product”, viewable from the firm, the trade or the consumer [3].

Aaker’s Managing Brand Equity (1991) sorted that added value into five asset categories: brand loyalty, name awareness, perceived quality, brand associations, and other proprietary assets such as patents and channel relationships [1]. Keller (1993) moved the unit of analysis into the customer’s head. His definition, quoted once here, is “the differential effect of brand knowledge on consumer response to the marketing of the brand” [2, p. 2]. A brand has positive equity when consumers react more favourably to a marketing action for the named brand than to the same action for an unnamed version [2, p. 2]. Brand knowledge is awareness plus image, and image is a set of “favorable, strong, and unique brand associations” [2, p. 1].

Kotler’s Marketing Management carried the Keller model into the standard curriculum from the 12th edition (2006), with Keller as co-author [8].

What the evidence shows

Perceived quality predicts stock returns. Aaker and Jacobson (1994) asked whether a survey measure of perceived quality (Total Research’s EquiTrend database) explained movements in firm value beyond what return on investment already explained. Controlling for economy-wide factors and ROI, they found “a positive relationship between stock return and changes in quality perceptions” [11, abstract]: the survey carried information the accounts did not.

Some survey pillars carry financial information, others do not. Mizik and Jacobson (2008) repeated the exercise with the five pillars of Young and Rubicam’s Brand Asset Valuator. Relevance and energy added information to accounting measures in explaining stock return; esteem and knowledge did not, because their effect was already in current-term accounting results [12, abstract].

Consumer-based and sales-based equity partly align. Datta, Ailawadi and van Heerde (2017) built the most direct test to date: ten years of IRI scanner data and BAV survey data for 290 brands in 25 packaged-goods categories. Sales-based brand equity (SBBE) is the brand intercept in a market-share model. The abstract reports “a fairly strong positive association of SBBE with three dimensions of CBBE, relevance, esteem, and knowledge, but a slight negative correspondence with the fourth dimension, energized differentiation” [13]. A published summary gives the correlations as 0.35, 0.39, 0.53 and -0.14 respectively [14].

Brands high in knowledge, esteem and relevance gained more from price cuts and distribution gains; differentiated brands showed a larger advertising response; and knowledge did not buy a price premium [15]. Ailawadi: “Academics as well as managers often regard a brand’s ability to charge a price premium as the single most important indicator of brand equity, but we find that it is not” [15].

Brand building reduces price sensitivity, slowly. Binet and Field’s meta-analysis of the IPA Databank (996 campaigns, 700 brands, 83 categories) reports the share of campaigns with a very large reduction in price sensitivity by the number of very large brand effects the same campaign recorded: 1% with none, 3% with one, 5% with two, 8% with three, 11% with four or more [21, slide 7]. Pricing effects also rise with campaign duration up to three years [21, slide 5]. Binet in 2023: “There are no examples of short-term activation campaigns lowering the price sensitivity of a brand” [22].

0 brand effects1%

13%

25%

38%

4 or more11%

Pricing power in the IPA data is a downstream effect of accumulated brand effects, which is the strongest published support for treating equity as a long-term asset and the clearest warning against expecting it quickly.

Source: Binet and Field, IPA, The long and the short of it: 10 key principles of success, 2013, IPA Databank meta-analysis, n=996 campaigns, 700 brands, 83 categories, slide 7. Redrawn by Brave New Media.

Unit: % of campaigns

Show the data

Item	% of campaigns
0 brand effects	1
1	3
2	5
3	8
4 or more	11

Cialdini did not address brand equity as a construct, but three of his principles describe how Aaker’s associations get made: social proof (chapter 4), authority (chapter 5) and commitment and consistency (chapter 7) of Influence, 2021 edition [23]. That is a claim about mechanism, not a finding about equity.

Buyers say trust decides purchases at the level of quality and price. Edelman’s 2025 brand trust survey, 15,042 people in 15 markets, asked which attributes of a brand are important or a deal breaker when buying. “I trust it” scored 88%, level with best quality and value for money, both 88% [30, p. 14]. The 2026 wave repeated the tie at 88%, 89% and 88% [31, p. 12]. That is stated intention, the survey side of the ledger, not the behaviour side.

Offers the best quality88%

Good value for the money88%

I trust it88%

High quality customer service85%

Has a good reputation85%

Convenient to find, buy and use84%

I love it82%

I trust the company that owns the brand80%

Trust is a stated purchase criterion at the same level as quality and price, and the 2026 wave repeats the tie at 88, 89 and 88. It is a survey of intention, not a measure of behaviour, which is exactly the gap between Keller's definition and a tracker score.

Source: Edelman, 2025 Edelman Trust Barometer Special Report: Brand Trust, From We to Me, 2025, n=15,042, 15 markets, question asked to half the sample, fieldwork 24 April to 5 May 2025, p. 14. Redrawn by Brave New Media.

Unit: % who say important or a deal breaker

Show the data

Item	% who say important or a deal breaker
Offers the best quality	88
Good value for the money	88
I trust it	88
High quality customer service	85
Has a good reputation	85
Convenient to find, buy and use	84
I love it	82
I trust the company that owns the brand	80

Industry valuers report large brand premiums, as associations. McKinsey’s B2B branding analysis has strong brands outperforming weak ones by 20% [33]. Brand Finance puts the 300 most valuable B2B brands at $4 trillion, 11% of their owners’ enterprise value, and reports a 65% premium in forward price-to-earnings ratios for the stronger-branded businesses [32]. Across its Global 500, brand value averages around 20% of enterprise value [26]. None of these separates the brand from the business that carries it, which is the problem the next section takes up.

The challenge

The three challenges do not say brands add no value. They say the number on your dashboard may not be what you think it is.

The word means three things. Feldwick (1996) sorted the uses into (a) the total value of a brand as a separable asset when sold or put on a balance sheet, (b) the strength of consumers’ attachment to a brand, and (c) the associations and beliefs the consumer holds about it [5, as summarised in 6]: brand valuation, brand strength, brand description. The commercial valuation systems mix all three. Interbrand starts from economic profit, multiplies by a Role of Brand Index, and discounts at a rate set by a 0 to 100 Brand Strength score across ten factors [9]. Kantar BrandZ starts from parent-company earnings, isolates a branded intangible portion, and multiplies by a Brand Contribution score from its Meaningful, Different, Salient survey [10]. Both are financial value times a survey multiplier. Brand Finance’s Global 500 is a third such system: $10.4 trillion across 500 brands in 2026, up 11% in a year, with the top ten brands holding 27% of the total [26]. None of the three is Keller’s differential effect. Interbrand’s Best Global Brands also requires a Brand Strength Score of 50 or above for entry [9]. A league table with an entry gate is a marketing artefact, not a measurement. The tables are also volatile in a way beliefs are not: Kantar’s Top 100 total fell 20% in 2023 and rose 29% and 22% in the two years after, a pattern Kantar itself attributes to the stock market [24].

The Top 100 was worth $1.44 trillion in 2006 and $13.1 trillion in 2026. Kantar describes the 2023 drop as a stock market-induced swoon; nothing consumers believed about brands changed by a fifth in a year, which is what a valuation built on earnings will do.

Source: Kantar BrandZ Most Valuable Global Brands, annual ranking pages and press releases 2017 to 2026; 2026 ranking built on 4.6 million respondents across more than 22,000 brands. Redrawn by Brave New Media.

Unit: US$ trillion

Show the data

Year	US$ trillion
2017	3.64
2018	4.4
2019	4.7
2020	5.0
2021	7.1
2022	8.7
2023	6.9
2024	8.3
2025	10.7
2026	13.1

Attitudinal equity may be restating market share. This is the Ehrenberg-Bass position. Double jeopardy, first documented by McPhee in 1963 and reviewed by Ehrenberg, Goodhardt and Barwise (1990), says a smaller brand has fewer buyers and those buyers buy it slightly less often [16]. The Institute’s restatement: “Double Jeopardy tells us that loyalty is largely a predictable function of market share” [17]. Sharp draws the consequence for surveys: “Larger brands tend to score higher on any image question, because they will have more customers among the survey respondents” [18, quoted in 28]. On the Institute’s site: “attitudes follow behaviour” and “the effect of behaviour on attitudes is much stronger” than the reverse [19]. His sharpest line targets the industry built on Aaker’s first asset: “Brand equity consultancies make money flattering owners of large market share brands… that they are not just large but also strong with attitudinally committed consumers” [18, quoted in 27]. Sharp does not deny loyalty exists; he calls it “a natural part of buying behaviour” [19, citing How Brands Grow p. 92]. He denies it is an independent asset you can grow ahead of penetration. Kantar’s Blueprint for Brand Growth, drawn from 6.5 billion BrandZ and Worldpanel data points, reports that brands that are meaningfully different to more people have five times the market penetration [29]. Read from the Ehrenberg-Bass side, that is the same correlation with the arrow reversed.

Market penetration5x

Volume share9x

Average selling price2x

Likelihood of growing share4x

The penetration multiple is what the vendor sells; Ehrenberg-Bass reads the same figure the other way, as large brands scoring higher on any survey because more respondents buy them. The volume, price and growth multiples apply to brands whose predisposition work is, in Kantar's phrase, optimally executed.

Source: Kantar, The Blueprint for Brand Growth, press release 14 May 2024, analysis of more than 6.5 billion attitudinal and shopper data points from Kantar BrandZ and Worldpanel. Redrawn by Brave New Media.

Unit: multiple (x); label each bar "5x", "9x", "2x", "4x"

Show the data

Item	multiple (x)
Market penetration	5
Volume share	9
Average selling price	2
Likelihood of growing share	4

Sharp and Romaniuk replace the asset list with two properties. Mental availability “means that buyers easily think of the brand when buying and recognize it quickly”; physical availability “means that buyers can buy the brand in a variant, quantity and price that matches their purchasing needs” [20]. Against Aaker, this keeps awareness and associations, drops perceived quality as a driver, and demotes loyalty from cause to consequence.

The tests are partial. Datta and colleagues found alignment for three pillars, not four, and it was moderate. Mizik and Jacobson found two pillars informative, not five. Aaker and Jacobson found one measure informative among large corporate brands. Datta’s team describe association, not cause [13]. Anyone who tells you a tracker score predicts revenue is ahead of the published evidence.

Where practitioners get it wrong

Reading a valuation ranking as a management score. The Interbrand and Kantar numbers are financial value times a multiplier. A brand that grows revenue with unchanged perceptions climbs the table; one whose company reports lower profit falls, whatever its buyers think. Feldwick’s first meaning is being sold as his second [5, 6]. Apple’s brand was worth $470.9 billion to Interbrand in October 2025, $608 billion to Brand Finance in January 2026 and $1.38 trillion to Kantar in May 2026 [25, 26, 24]; Kantar ranks Google above Apple, the other two do not.

Interbrand 2025Brand Finance 2026Kantar BrandZ 2026

Apple470.96081380.3

Microsoft388.5565.21111.8

Google317.14331484.9

Amazon319.93701022.8

Three firms, three methods, one brand, and values that differ by a factor of three. The order also disagrees: Kantar puts Google first, Interbrand and Brand Finance put Apple first. Whatever each number measures, it is not the same thing, and none of them is Keller's differential effect.

Source: Interbrand, Best Global Brands 2025; Brand Finance, Global 500 2026; Kantar BrandZ Most Valuable Global Brands 2026 (4.6 million respondents, 22,000+ brands). Each firm's published figure. Redrawn by Brave New Media.

Unit: US$ billion

Show the data

Brand	Interbrand 2025	Brand Finance 2026	Kantar BrandZ 2026
Apple	470.9	608	1380.3
Microsoft	388.5	565.2	1111.8
Google	317.1	433	1484.9
Amazon	319.9	370	1022.8

Treating tracker “loyalty” and “favourability” as assets to grow. Double jeopardy predicts most of the variation in these metrics from share alone [16, 17]. The useful diagnostic is the deviation from the double jeopardy line, not the raw score. Self-report is no better: The CMO Survey asks marketing leaders to rate the change in their own brand value, and the February 2025 wave averaged +8.4% across 167 answers, with no instrument behind the number [34, p. 43].

Assuming equity buys a price premium on its own. Datta’s team found high knowledge did not support a premium in packaged goods [15]. Binet and Field found price effects arrive only with multiple large brand effects and only over years [21]. Raising price the quarter after a good tracker wave is a bet the evidence does not back.

Our own observation. Five projects, one developer behind them, and the developer brand was the least communicated asset in the group. Fifteen years, thousands of apartments delivered, energy efficiency as standard: none of it was being said. We gave the company its own campaign, separate from any project, because in a cautious market the most useful thing a developer can prove is that it will still be there at completion. We cannot yet show that equity in a price. Enquiries are not reservations, and the link to signed contracts sits in the developer’s sales data, still to be joined.

What to do with it

1. Write down which of Feldwick’s three meanings you are measuring before you buy a tracker or a valuation [5]. If the answer is “all three”, you have not decided.
2. Estimate sales-based equity from your own data: the brand intercept in a share or choice model after controlling for price, promotion, distribution and advertising [13]. Failing that, use revenue premium over a private-label equivalent.
3. Plot every attitudinal metric against penetration for your category and report the residual, not the level [16, 17].
4. Track awareness and associations as Keller specified them, favourable, strong and unique, by category entry point [2, 20]. Drop indices that average across them.
5. Put a behavioural measure next to every survey measure on the same page: penetration, repeat rate, distribution, price paid [12].
6. If the objective is pricing power, plan for three years or more and for brand-building that produces several measurable brand effects: that is the only pattern in the IPA data associated with large falls in price sensitivity [21, 22].
7. Retire the phrase “grow loyalty” unless your loyalty metric sits below the double jeopardy line for your share [17]. Otherwise the plan is “grow penetration” under another name.