Working on the assumption that people of a similar age and social class exhibit similar behaviour, marketers use demographics as a shorthand - young/old, upmarket/downmarket.
I'm not the first to point out that this is a gross over-simplification. But demographic information is an aggregated view that's intended to describe group characteristics, it's not intended to predict an individual's behaviour. While knowing that one group is more likely to buy our product than another allows us to target them, thereby increasing our chances of success, we recognise that there will be a substantial number in that group who won't respond to the message for one reason or another.
We've come to live with this compromise - and entire media planning and trading systems have grown up around it. But there is a better way, and the biggest exponent of it, Google, already makes more money in the UK than ITV.
Google's success is based on relevance. Advertising is bought against keywords - terms that users search for - and, of course, these keywords describe the interests of the searcher. Each keyword has a bid price that depends entirely on demand, and the whole system is automated from Google's side.
The genius of all this is that consumers get ads that are directly relevant to their search, presented to them exactly when they've just indicated their interest in a topic. And the content costs Google nothing.
It's not surprising, then, that when Facebook launched an online advertising system, it looked to the search engine for inspiration.
Its system imports features that made Google successful - self-operation, combined with a credit-card payment facility that makes it easy for small businesses to participate. An auction allows demand to influence pricing, and you can target users' interests based on the groups they belong to.
But it has gone much further than this. Social ads can be served to friends of users. If I buy a book on your site (assuming I've arrived via Facebook's beacon system), my friends can be shown an ad - 'Andrew Walmsley rated this book 4/5 - buy it here' along with my picture to emphasise the personal connection.
And whereas only individuals could set up pages before, Facebook now allows companies to do so, providing a base for promoting their products and services. Reflecting this, users don't 'friend' these pages, they become 'fans' - an important distinction.
So what has Facebook created? Social media has challenged marketers, because there isn't a clear role for brands in the space. For many consumers, brand presence feels like an intrusion into personal space, so success in this area has principally been limited to entertainment brands.
Facebook has achieved three things. It has found a distinctive way of creating a place for businesses to co-exist with people in a social network. It has established a means of promoting brands that's derived from people's interactions (and transactions). And it has created a clever way of targeting people that reflects their interests and behaviour.
The great thing about online media space is that nobody can be quite sure what use it is going to be to anyone. Facebook has given us a big box of toys, and it's up to us to figure out how to use it.
Google and Facebook give us a glimpse of the future - one where demographics will no longer be the means by which we understand audiences nor the currency by which we trade them. There's no place here for the compromise that demographics force on us, and it'll change both advertisers' and consumers' expectations across the media world.
- Andrew Walmsley is co-founder of i-Level
30 SECONDS ON ... RELEVANCE
- The Compact Oxford English dictionary says relevance is 'closely connected or appropriate to the matter in hand'.
- In 1986 cognitive scientist Dan Sperber and linguist Deirdre Wilson produced a paper on 'relevance decisions' in reasoning and communication. It stated: 'Any utterance addressed to someone automatically conveys the presumption of its own relevance. This fact, we call the principle of relevance.'
- In his Treatise on Probability published in 1921, economist John Maynard Keynes defined relevance in relation to risk calculations. He suggested that the relevance of a piece of information should be defined by the changes it produces in estimations of the probability of future events.
- In computer science, relevance is a score assigned to a search result based on how well it meets the information need of the searcher.