Your SEM matters to marketers
Its replacement, the “next level” Socio-Economic Measure (SEM), is said to be a more accurate macro-economic standard against which to measure how people live, based on what they have access to.
SEM has become an important tool in marketers’ arsenal to help them understand target audiences’ income, lifestyle, geography, spending habits and other factors.
Earlier this week, the Publisher Research Council (PRC) and Broadcast Research Council announced the opening up of access to their SEM weightings to provide media planners with a more stable picture of a household’s status. The open-source measure is available as a SEM calculator on the PRC website, while the Broadcast Research Council (BRC) is hosting a dropdown menu featuring a downloadable SEM user guide on its site.
The LSM, popular in the 1980s and 1990s, has largely fallen out of favour due to its racial bias and inaccurate measurement.
For decades, communication strategies were built around high LSMs and the aspirational classes, but it’s been described as “fatally flawed”.
Gary Whitaker, the chief executive of the BRC, says the LSM might have worked once, but it’s inappropriate now: “The LSM reflected a massive middle class, which, if you look at how many people pay tax and how many are unemployed, doesn’t make sense in our context. It also shows a perfect distribution of wealth. For example, if you present an LSM questionnaire to poorer people, who live communally, they might be asked: ‘Do you have a TV?’, and they would say yes, because they have access to a television, which would mean they would rate far higher on the LSM than they actually are.”
The LSM factors in 29 variables, which are heavily dependent on media consumption and durable goods, such as where people live; their type of residence; access to piped tap water and a flush toilet inside their homes; whether or not they have hot running water, a domestic worker or gardener and home security; the number of cellphones in the household; television sets; tumble dryer; pay-television; and computers.
And because the median age of white people is 39 and blacks is 24, LSMs are inherently biased, because it takes about 15 to 20 years to accumulate these durables, giving white people a head start and skewing advertising towards that market.
By contrast, SEMs factor in 14 variables or inputs, such as whether or not there’s a police station and post office near your home, the existence of household items such as a built-in kitchen sink, microwave, washing machine, floor polisher/vacuum cleaner, a free-standing deep freezer or side-by-side fridge and freezer, type of flooring, toilet types and location, roofing type, number of bedrooms, as well as motor car and home security service.
Each person is scored between 0 and 100.
But, Whitaker says, the LSM is a language that marketers understand: “They can tell you what LSM 1 to 10 means. When SEMs came about, every effort was made to educate marketers, but there were barriers to entry. For anyone to use SEMs for their own studies, they had to pay licence fees (to access the weightings), but that requirement is now lifted. It was a major barrier before. A lot of industry bodies are taking the SEM algorithms into their studies, so people are becoming more comfortable with using it.”
Peter Langschmidt, research consultant at the PRC, says the LSM shows an “almost perfect” bell-shaped population distribution, which suggests a massive middle class and income distribution similar to Australia and Canada - a far cry from our small middle class.
And South Africa’s Gini coefficient, a measure of inequality, is one of the highest in the world.
Langschmidt says LSMs were perfect in the late 1980s, when the minority of (white) households had electric stoves, microwaves, fridges, TVs and washing machines. Now that the majority of people possess these durables, this measure is meaningless.
He believes that although well-intentioned, LSMs, which were intended to be aracial at the time, have come back to bite the industry. “It’s completely contrived. We didn’t update LSMs for years at a time. LSMs created fairy tales of the middle class, but 40% of our people are poor. When you do brand profiles, you need a true consumer typology.”
SEMs, in contrast, allow for a more precise targeting of the market, and less wastage of advertising spend. “To define it, we created supergroups, which were natural breaks in segments: grouping society in poor, average and above average. The top end earns over R43000 a month, while the bottom end earns R4400.”
SEM is also more user-friendly, culturally relevant and an improved approach to segmentation.