We Have To Learn How To Read, Write, and Speak in Big Data

In his modern fable “The Library of Babel,” Argentinian writer Jorge Luis Borges describes an infinite library comprised of hexagonal rooms in which people are born, grow up, and grow old.

Each side of each room in the library contains five shelves fitted neatly with 35 books apiece, with each book containing exactly 410 pages, and each page of every book consisting of precisely 40 lines of script. As far as the books go, only a few words and phrases here and there make any actual, grammatical sense to those who try and read them. As the unknown narrator of Borges’ story — a man who has spent his entire lifetime within the confines of the Library — puts it, “This much is already known [about the books]: for every sensible line of straightforward statement, there are leagues of senseless cacophonies, verbal jumbles and incoherences.”

In our own age of big data, stories like Borges’ “The Library of Babel” start taking on new resonance. While not quite infinite — not quite yet — big data does seem intent on moving in that direction. Since the commercial Internet began some 20+ years ago, the amount of data contained in our world now doubles every two years. According to the MIT Technology Review, only 0.5 percent of that data has ever been analyzed or “read” in any shape, form, or fashion.

Unlike any other era in human history, we live in an overwhelming surfeit of knowledge that accumulates by the megabyte every second. Like the group of human beings marooned in Borges’ strange, unending maze of books, we find ourselves confronted with the daily task of sorting through zetabytes and zetabytes of “verbal jumbles and incoherences” in order to arrive at a few legible, logical correlations.

I’m looking at you, marketers.

For advertisers and agencies that hope to remain competitive in the early 21st century, new skillsets are required beyond simple creativity and shrewd business acumen (though it’s not like those skills aren’t sorely needed, also). Advertisers need to learn how to “read” data points at the log level, meaning they need to learn how to interpret and draw actionable insight from a fast-developing stream of expansive data points — data that’s actively being “written” by individual consumers in near-real time even as those consumers browse the internet, or use their mobile apps, or watch their OTT television screens.

By accumulating enough log-level data points, and by carefully “connecting the dots” between these log-level data points (and even their own proprietary data), advertisers can suddenly “spell out” correlations between peoples’ online actions that — just a moment ago — seemed utterly random and impossible to interpret. By attaining that granular level of data clarity, advertisers find themselves in a position where they can serve ads that “speak” to consumers like the individuals they are; ads that they not only understand but also are able to “click with” personally; ads they find relevant; that they can relate to; ads that are able to “speak their language”.

Marketers who are able to understand this expansive stream of data points will be able to speak to customers at the individual level. They’ll learn to communicate to customers in ways where other advertisers — those who haven’t brushed up on their log-level data science skills — are bound to fail. And by so doing, they’ll have achieved a form of language that’s truly unprecedented, certainly in advertising.

Of course, being fully “fluent” in big data will remain an impossible challenge, since data will continue to pile faster than any single human or machine intelligence can ever possibly process and/or “shovel” it. And while the rate of interpretable data might someday rise beyond a mere 0.5 percent, that’s an unreasonable goal, at least for the time being. But that’s no reason for marketers not to try and become as “log-level literate” as possible. It’s not as if big data’s been around since time eternal. We’ve only begun to read, write, and speak it.

In order to un-maze ourselves of big data; to free ourselves from an over-saturated, glutted world of information; and to read freely from the vast, near infinite library that surrounds us — we, as marketers, need to acquire all the skills of data scientists. Only by doing this will we begin to make sense of big data — and use it sensibly with our audiences.

We need to be able to monitor the log-level flow of datasets and develop correlations between different data points that we have yet to fully make sense of. To borrow from Borges’ story, we need to be willing to “think outside the hexagon,” and be willing to look at larger data samples gathered from additional rooms.. We need to learn how to build our own cuneiform by which to read and write and communicate with our customers by. Not because there’s something noble or idyllic about the prospect of being able to “read” and “write” in log-level data. But because it’s commonsense and because it translates as “good for business”.

But in order to engage consumers meaningfully in the programmable era of marketing, we first need to learn how to speak their language. And the language that digital consumers speak is none other than log-level data.