By the Numbers

The brands and companies listed in the Superbrands report represent a huge chunk of the billions spent on advertising in the U.S. each year. As in past years, the individual brands are ranked in two ways:

Brands listed in the 15 category sections are ranked among their peers according to full-year 2008 sales in the U.S. market (unless otherwise noted). Here, readers will find competitive analysis and industry trends compiled by the Brandweek staff along with ad agency affiliation, full-year media spending and the brand's rankings among consumers, per the Harris Interactive EquiTrend survey (see "Study Key" below).

Brands in the Top 2000 tables are ranked by full-year 2008 measured media expenditures in the U.S., as tallied by Nielsen Monitor-Plus.

That said, to be considered a Superbrand, a particular nameplate must be advertised in a quantifiable manner across Nielsen Monitor-Plus' media classifications (TV, print, outdoor, radio, etc.). The list is biased against corporate trademarks. Brands that engage in under-the-radar marketing as a predominant tactic aren't likely to appear in these tables.

Consumer perceptions of brands from the EquiTrend study by Harris Interactive, Rochester, N.Y., round out the package. Comprehensive diagnostic results for evaluating brand equity begin with "Quality" and "Familiarity," a sensitive measure of a consumer's ability to rate a brand. Respondents then ranked brands with which they were "somewhat," "very" or "extremely" familiar on their propensity to purchase that brand to better gauge its vitality. Purchase intent correlates strongly to loyalty.

A sample of over 20,000 consumers ages 15-and-over were surveyed online. The interviewees came from Harris' online panel of respondents, a multimillion-member database that has agreed to take part in online surveys. Each respondent was asked to rate a total of 60 randomly selected brands. Each brand received approximately 1,000 ratings.

Brands were randomized so names did not appear in the same order. Data was weighted to be representative of the U.S. population on the basis of age, sex, education, race/ethnicity, region and income.

Equitrend Study Key: Each brand was rated on the following criteria:

Familiarity, 0-100%: Respondents were asked to rate familiarity with each brand on a 1 to 5 scale, where "1" means they never heard of the brand, "2" means they just know it, "3" means they are somewhat familiar with it, "4" means very familiar and "5" means they are extremely familiar with the brand. Those results were then adapted to a percentage rating. For example, if 800 people out of 1,000 were between 3 and 5, then the brand would receive an 80% rating.

Quality, 0-10: The cornerstone of the EquiTrend brand measurement system is the measurement of perceived quality, where zero means "Unacceptable/poor," 5 means "Quite acceptable" and 10 means "Outstanding/Extraordinary." Respondents could rate a brand using any number in that range, or indicate that they had absolutely no opinion about the brand.

Purchase Consideration, 0-10: Conceptualized as a measure of the extent to which a consumer intends to have a future relationship with a brand, respondents indicated their answers using a 1-4 scale, where "1" means "never," "2" means "not likely," "3" means "possibly" and "4" means "absolutely" would purchase a product brand/watch a TV channel/show. For comparibility, Purchase Intent was rescaled to the same 0-10 score as Quality.

Equity, 0-100: A brand's equity score is determined by a calculation of Familiarity, Quality and Purchase Intent, as a way to further our understanding of a brand's overall strength. Brands that have high quality, are well known and score well on purchase intent, have greater equity than brands that are not well known, have poor perceived quality or those brands which consumers are not interested in purchasing. The actual equation is done by indexing Familiarity and Purchase Intent; followed by weighting the Familiarity score*, which is then multiplied by the mean of Quality and Purchase Intent, with the result indexed on 100.

*Familiarity weights: 5 = 1; 4 = 0.9; 3 = 0.8