Andy Ho Senior Writer

Imagine matching politician to voter - now that's a brainwave

Market research firm Nielsen says that many advertisers in Singapore are relying on neuromarketing to pitch their products to consumers.

This is the use of neuroscience to match products with people, where science tries to determine how the human mind processes information. The idea is to look for which brain areas are related to what thinking processes. And advertisers use this knowledge to see how best they may convince consumers to buy a product or service.

The most often cited study was that published in 2004 in which brain scans were done on people as they drank Pepsi or Coca-Cola.

If they were told what they were drinking, an area of the brain's frontal lobe fired up and they chose Coke over Pepsi. But if they were not told what they were drinking, an area of the limbic system fired up and most felt that Pepsi tasted better than Coke.

Since what these two brain areas actually do is quite established, it was concluded that Coke prevailed in the consumer's subconscious whereas Pepsi prevailed if one had tasted it previously. So brand image (in the subconscious) could affect customer behaviour more than product characteristics (like taste) themselves.

According to Nielsen, the demand for such services has tripled in the last two years. To track brainwave activity, Nielsen uses a decades-old technology called electroencephalography, which is integrated with a technology to track eye movements to study consumer brain activity.

Overseas, though, neuromarketing usually also involves the use of scans, like those used in the Cola study, called functional magnetic resonance imaging (fMRI).

What is fMRI? If you have had a headache for weeks, your doctor may order a standard MRI for it. This images the brain's structure to look for structural problems like, say, a tumour. But it cannot indicate how the relevant brain areas may be functioning. By contrast, fMRI can identify how specific brain areas are functioning. In the fMRI scan, the subject is asked to think about, look at or listen to something. For example, a proposed TV ad is shown to the test subject as he lies inside the scanner. These tasks cause increased activity in the brain area responsible for the thought process that is evoked by the specific stimuli.

What fMRI actually looks for are changes in blood flow to those areas - specifically, whether the blood that flows to those areas has lots of oxygen or very little of it. This data is used to produce pictures of the brain that are divided into little digital cubes called voxels. Sophisticated statistical software will count the number of voxels with oxygen- rich and oxygen-poor blood respectively and then calculate their ratio.

It is assumed that the brain areas where neurons have just fired off will receive an inflow of oxygen- rich blood just a few seconds afterwards. So the ratio of oxygen-rich to oxygen-poor voxels should show the brain areas that were most active during a particular mental task.

This data is then transformed into pictures of the functioning brain, which marketing research uses in an attempt to uncover hidden information about how consumers make their choices.

To fight declining sales, Campbell used fMRI studies that showed the label and packaging of its canned soup were turning consumers off. These were then redesigned to evoke feelings of warmth. Launched in 2010, the new look saw sales pick up.

Some of such information cannot be obtained using conventional methods like focus groups or product tests in the market because consumers may not be able to articulate cogently what their preferences are if asked directly.

However, if their brains held that hidden information about their true preferences, imaging the brain could show companies not only what consumers like but also what they will actually buy.

Consumers are not likely to be able to tell that an advertiser is getting to them through this new way. Their main line of defence lies with practitioners restraining themselves. For example, the US Neuromarketing Science and Business Association last year drew up a Code of Ethics for its members; it promises not to manipulate study subjects and consumers into unhealthy consumer habits.

Now apply this whole neuromarketing idea to the citizen not as consumer but as voter. If fMRI can tell in which brain area a specific process occurs, it could reveal how political judgment and decisions, including voting, are related to what brain areas and processes in voters.

If so, political parties might one day be able to have their messages and candidates tailored for different segments of the voting public accordingly.

While it is not known if anyone anywhere is making use of neuromarketing to select candidates or craft their messages, the political possibilities have been studied since 2006, when the first bona fide empirical studies in this new field, using fMRI to examine voting preferences, were published in good journals.

Political parties presumably groom their potential candidates for public office via their tried and tested ways. But fMRI could be brought to bear on the process. For example, a candidate whose face is perceived by voters to be worthy of trust is more likely to win votes. A lot of fMRI work on how the human face is perceived, including its colour and symmetry, has already been done. Which brain structures are involved in processing which parts of the face is known too.

A recent study found that if a candidate's face activates the brain area called the insula, he is more likely to lose in an election.

Then there are different dimensions of trust, including reputation and fairness, which have been identified as activating specific brain areas on fMRI. Drugs like oxytocin have been used experimentally in studies to increase trust and fMRI changes can indeed be detected in these areas.

Likewise, when we anticipate something pleasant - like tasty food, nice music, pretty girls or monetary rewards - part of the orbitofrontal cortex is activated on fMRI. Perhaps the content of and nuances in a (facially likeable) candidate's message can be tweaked to maximise responses in that particular brain area.

A 2013 study even identified mental activity in specific brain areas that were associated with conservatism which were clearly different from the brain areas associated with liberalism.

The study concluded that "liberals and conservatives differ in terms of brain responses to stimuli, brain function and even brain anatomy". So fMRI could identify which brain activity correlates with specific political ideologies. This is quite an achievement for such a young sub-discipline.

Using fMRI to find correlations between the activation of specific brain areas and political attitudes is arguably less hampered by the subjective biases that implicate self-reporting in opinion surveys, for example. Political parties with deep pockets may, in short order, resort to neuromarketing to match politician to voter. If this succeeds, fMRI could change far more than consumer behaviour.

andyho@sph.com.sg

Join ST's Telegram channel and get the latest breaking news delivered to you.