The SAARF LSM and the defining power of fridges

The simple answer is that we're not defining people by their fridges - in fact, we're not even defining their standard of living by just their fridges. The SAARF LSM is a multivariate discriminator consisting of an ever-changing list of variables that keep track of a dynamic South African society.

Every year, SAARF tests dozens of variables to determine which are the best predictors of what makes people similar to those within their LSM group, and different to those in other LSM groups. What makes an LSM 4 an LSM 4, and what makes this LSM 4 different to an LSM 5, and hugely different to an LSM 10?

Currently, 29 variables are used, and they change over time, as variables like having a sewing machine, are removed, and replaced by stronger living standard predictors such as having a swimming pool.

No one variable can make or break someone's LSM ranking. A common misconception for instance, is that having a pay-TV subscription automatically makes you an upper LSM. You can be a premium DStv subscriber, but without checking the boxes next to a whole basket of other variables, you could still be in LSM 4. It takes a lot more than one variable to have an impact on your LSM status.

LSMs in action

Let's take the fictional Thabo as an example. He lives in a shack in the Zandspruit informal settlement. He has no modcons to make his day-to-day existence easier: no washing machine, no fridge, no microwave oven, and a radio's his only access to media. He must walk down the road to fetch water and use a toilet. He is in LSM 2.

He wins a spot of cash in the lottery. Since he's a total soccer fanatic, he uses the windfall to buy a TV and get himself a DStv subscription for SuperSport. Do they make a difference to his standard of living? Certainly. But they don't rocket him into LSM 8. They bump him up to LSM 3 - but only just.

His luck holds out and he finds a job. He buys a fridge and a microwave oven on HP. This moves him up to LSM 4. He moves to a shack with a tap on the plot and an outside loo, and he buys himself a top-loader. Up he goes to LSM 5. At this point, Thabo's standard of living is far better than it was before, but he'll still need to acquire a great deal more labour-saving devices, move into a house, buy a car and install a geyser, to name but a few, before he can move into the top LSMs.

Knowing a person's living standard gives you insight into how they might behave. Having a domestic worker for example, means you can come home and watch TV or read a magazine, where other people might have to spend that time doing household chores. Having a fridge changes your shopping habits - you can't bulk-buy perishables without a fridge or freezer. And so it is with each of the variables that are included in the LSM...there are good reasons for their inclusion.

Using the right tool

But knowing someone's living standard does not always help you. Sometimes, using the LSM isn't the right segmentation tool for the job. Maybe you're comparing Fairlady with Cosmo - you'll find the LSM profile is very similar for both titles, making the LSM of little value in this situation.

This is because no segmentation tool is the alpha and the omega, not even the LSM which so many marketers use to the exclusion of all else. This is why SAARF has three other segmentation tools besides the LSM, to segment markets along attitudinal, lifestyle, and lifestage lines, and there are other more qualitative segmentation tools from other providers besides.

Segmenting a market isn't an either/or situation. You're allowed to - in fact, you're encouraged to - use more than one segmentation tool. And choose the best tool for the job. Just like you wouldn't use a ruler to measure the length of a rugby field, you should use the right segmentation tool depending on what you want to measure.

If your gut tells you that age would be a better differentiator, as it is with the example of Fairlady and Cosmo, then use age. If you suspect it's people's opinions or attitudes or mindsets driving behaviour, then use another tool, like SAARF Attitudes.

Or make your own segmentation tool. Add AMPS variables, like household income*, gender, or working status, to your LSM run.

To sum up, the LSM is an extremely powerful segmentation tool because its basis is multivariate statistics: it doesn't rely on only one variable. It provides a totally objective segmentation, but remember that it can't always give you all the answers you need.

It's all about choosing the right tool for the job. After all, you wouldn't use your fridge if you wanted to fry an egg, would you?

*Many people think that 'LSM' is a synonym for 'income'. It is not, although there is a correlation between income and LSM (but just because someone's in LSM 10, doesn't mean she earns a huge salary). Income is not one of the variables making up the LSM.