Data Is The New Snake Oil?

You can't log into a business related social media platform these days without tripping over the claim that "data is the new oil", but what sort of oil are we talking about here? The claim is made on the basis that crude oil (of the sticky hydrocarbon variety) is of much less value than its refined and productized equivalents, and this is paralleled with the vast quantity of cheap and not-so-cheerful data that is likely to come gushing from everywhere in the internet of everything, which also has little or no value before intelligible information has been derived from it. So analytics (it is hoped) will release so much latent value from the tidal waves of data that are about to be unleashed by things like the IoT that the software and related services industries will be lighting their cigars with thousand dollar bills for the coming century.

Whilst I do not doubt that money will be made and insights (some at least) will be gained, I have to remain a little skeptical of the analogy. For a start, on perhaps a somewhat pedantic note, crude oil is crude oil. A 42 gallon barrel of crude contains probably something around 5.8million Btu. It takes maybe 75Btu to make a cup of coffee, and so that barrel can make around 77,000 cups - enough to keep me caffeinated for about fifty three years.

Of course there are various grades of crude, and the quality of some can be pretty poor, requiring a lot more processing, but even so, there are years of cups of tea and coffee in every barrel. A megabit of data from a useless sensor is worth the same as a gigabit of data from the same useless sensor…no Joe…

So let's remember that not all data is created equal, and whereas if your crude does not contain a combination of alkanes, cycloalkanes and aromatic hydrocarbons it just isn't crude, data is data, and the vast majority of it out there is not going to refine into anything particularly useable, no matter how smart Watson might get.

Secondly, let's just take a look at where the more 'useful' seeming data (the stuff about people) is going to come from.

Remember in the old days of the web, they used to say that 'content is king'? It sure used to be when people were the most common cowboys riding the internet range, but that hasn't been the case for some time now. Search optimised content may still be king, but king of what, precisely? Land of the bots and spiders, of course. The internet now consists of two related but separate parallel universes, one for the machines - full of words that people never read and images designed to drive traffic - alongside hooks to the useful downloads, videos and services that humans actually find of interest.

What makes anyone think that 'valuable' data in the future is going to be any different?

The real name, of a real person, with an actual email address, or the number of a cell phone that a real person might actually answer will of course be as hard to get your hands on as the proverbial unicorn poo (or we would like to think so anyway, if people have gotten their heads around the concept of privacy), but I (perhaps cynically) am not sure if that will even matter to the majority of analytics folk.

In the same way that web traffic figures alone mean very little these days but remain a metric the marketing team like and are willing to spend time and money improving, I get the feeling that bland and (partially) bogus insights electronically derived from the swirling lake of internet data will be just as frequently referred to in order to serve up business intelligence reports and somehow make money; and while 'personal profiling analytics' (the equivalent of SEO) tailor our web-personae to achieve whatever objectives we're trying to achieve by remaining visible on the web, 'personal information spiders' will scour the web and pick out our artificial selves to include in their business intelligence reports.

Something that can be converted into money is the new oil, where there is demand from a broad section of the population and a willingness to pay a consistent and sustainable amount of their weekly income to get that something in order to survive. That's not data, not all of it anyway.