What do consumers have in common with earthquakes? Both, researchers say, broadly conform to patterns described by a formula known as Zipf's law.
In the 1930s, the Harvard linguist George Kingsley Zipf found that "the" - the most-used English word - occurred about twice as often as "of"' (second place), about three times as often as "and" (third) and so on.
Since then, scientists have found similar relationships between the size and frequency of earthquakes and in a variety of other natural and artificial phenomena.
Jan Hofmeyr, an expert on consumer behavior at Synovate, the market research arm of the London advertising company Aegis, said he had recently discovered that Zipf's law also applied to the brand preferences of consumers and their spending habits.
Like the fictional Harvard professor Robert Langdon in "The Da Vinci Code," which featured religious mysteries and obscure mathematical clues, Hofmeyr, a South African, understands parallels between the sacred and the profane. He earned a doctorate in the psychology of religious belief formation before doing political research for the African National Congress.
His latest work, Hofmeyr said, has a more earthly purpose: helping brands get more out of their market research and their ad spending.
Market research, like other marketing disciplines, has been gripped by a demand for measurable results. No longer is it good enough simply to study who uses brand-name products and what they think about them. Clients want a formula for turning research into revenue.
"With this approach, the moment you determine a brand's ranking, you can predict the market share," Hofmeyr said. "Marketers have always known it's better to be No. 1 than No. 2, but now you can attach a revenue consequence to that."
Before being adopted by Synovate for use in a new research method, Hofmeyr's ideas were tested in Australia on two product categories: toilet paper and instant coffee. Consumers were asked to identify the brands they used and to rank them in order of preference.
According to his model, consumers who used four brands of toilet paper might have been expected to spend about 53 percent of their toilet paper budget on the top choice, 27 percent on the second brand, 13 percent on the third and 7 percent on the fourth.
In the Australian test case, consumers actually spent 50 percent on the top choice, 33 percent on the second, 9 percent on the third and 8 percent on the fourth. Averaged over hundreds of consumers, Hofmeyr said, the study showed an unusually high correlation between stated preferences and actual purchase decisions.
To be sure, market research theories and practices are proliferating like products on store shelves. Hofmeyr, for instance, previously helped develop another system called the Conversion Model for a rival business, TNS.
One research method currently in vogue involves comparing the number of people who say they are likely to recommend a product with those who would warn others not to buy the same brand. This calculation produces a "net promoter score," which is seen as a way to measure a brand's word-of-mouth performance. In the era of online social networks, where good or bad news about a brand can spread like wildfire, that concept is increasingly prized.
Other research methods involve asking consumers to rate brands on 10-point scales for factors like value, likability and satisfaction.
The problem with these approaches, Hofmeyr said, is that they sometimes look at individual brands in a vacuum. For instance, a brand that rates a 9 on a 10-point scale measuring product satisfaction might think it is sitting pretty - unless its main rival scores a 10.
Zipf's law clearly demonstrates the relative benefits of moving up in the rankings. A product that leaps from second to first in a category can really affect a company's bottom line. The advantage of moving up to, say, No. 5 from No. 6, is much smaller.
That knowledge, Hofmeyr said, could help marketers decide how best to spend their clients' money. Should they invest in corporate-image building, for example, or focus on more mundane issues like upgrading complaint-handling capabilities?
For market researchers, at least, that kind of information is the holy grail.
In the 1930s, the Harvard linguist George Kingsley Zipf found that "the" - the most-used English word - occurred about twice as often as "of"' (second place), about three times as often as "and" (third) and so on.
Since then, scientists have found similar relationships between the size and frequency of earthquakes and in a variety of other natural and artificial phenomena.
Jan Hofmeyr, an expert on consumer behavior at Synovate, the market research arm of the London advertising company Aegis, said he had recently discovered that Zipf's law also applied to the brand preferences of consumers and their spending habits.
Like the fictional Harvard professor Robert Langdon in "The Da Vinci Code," which featured religious mysteries and obscure mathematical clues, Hofmeyr, a South African, understands parallels between the sacred and the profane. He earned a doctorate in the psychology of religious belief formation before doing political research for the African National Congress.
His latest work, Hofmeyr said, has a more earthly purpose: helping brands get more out of their market research and their ad spending.
Market research, like other marketing disciplines, has been gripped by a demand for measurable results. No longer is it good enough simply to study who uses brand-name products and what they think about them. Clients want a formula for turning research into revenue.
"With this approach, the moment you determine a brand's ranking, you can predict the market share," Hofmeyr said. "Marketers have always known it's better to be No. 1 than No. 2, but now you can attach a revenue consequence to that."
Before being adopted by Synovate for use in a new research method, Hofmeyr's ideas were tested in Australia on two product categories: toilet paper and instant coffee. Consumers were asked to identify the brands they used and to rank them in order of preference.
According to his model, consumers who used four brands of toilet paper might have been expected to spend about 53 percent of their toilet paper budget on the top choice, 27 percent on the second brand, 13 percent on the third and 7 percent on the fourth.
In the Australian test case, consumers actually spent 50 percent on the top choice, 33 percent on the second, 9 percent on the third and 8 percent on the fourth. Averaged over hundreds of consumers, Hofmeyr said, the study showed an unusually high correlation between stated preferences and actual purchase decisions.
To be sure, market research theories and practices are proliferating like products on store shelves. Hofmeyr, for instance, previously helped develop another system called the Conversion Model for a rival business, TNS.
One research method currently in vogue involves comparing the number of people who say they are likely to recommend a product with those who would warn others not to buy the same brand. This calculation produces a "net promoter score," which is seen as a way to measure a brand's word-of-mouth performance. In the era of online social networks, where good or bad news about a brand can spread like wildfire, that concept is increasingly prized.
Other research methods involve asking consumers to rate brands on 10-point scales for factors like value, likability and satisfaction.
The problem with these approaches, Hofmeyr said, is that they sometimes look at individual brands in a vacuum. For instance, a brand that rates a 9 on a 10-point scale measuring product satisfaction might think it is sitting pretty - unless its main rival scores a 10.
Zipf's law clearly demonstrates the relative benefits of moving up in the rankings. A product that leaps from second to first in a category can really affect a company's bottom line. The advantage of moving up to, say, No. 5 from No. 6, is much smaller.
That knowledge, Hofmeyr said, could help marketers decide how best to spend their clients' money. Should they invest in corporate-image building, for example, or focus on more mundane issues like upgrading complaint-handling capabilities?
For market researchers, at least, that kind of information is the holy grail.