My digital shadow looks nothing like me

Data agencies gather and sell your information, even when it's not accurate.

I have a shadow. There's the Dan Cooper writing these words right now, standing at his desk in an attic in Norwich, England. There is also the Dan Cooper who has the same name and address but who only exists inside a computer sitting on a shelf. I had never heard of this man until a couple of months ago, but now I am intimately familiar with who he is, his contradictions and the terrible truth he may reflect upon me.

In the wake of Europe's new privacy laws, I polled numerous companies to learn what they knew about me. One of them was Acxiom, a principally American marketing-data agency that collects data on individuals across the globe. Data that it has bought, or gathered, is algorithmically mashed together with public records and drained through a series of statistical models. The eventual aim of this effort is to create a series of conclusions about user behavior that can be used to create increasingly targeted advertising.

Acxiom sent me a 24-page file that covered everything it had about me. That information has been sold to a number of "respected brands," including major names like Ford Motor Company, British Telecom and (British retailer) Tesco. The file was also passed on to companies that already know me pretty damn well: Facebook, eBay, Twitter and PayPal.

I'm sure that at some point in the past, I've unwittingly ticked "agree" on some box and given my consent for my data to be collected. And these companies are likely to have tried to build a complete profile of my health, economic status and purchasing profile. It's quite possible that the data has been used to send me specific offers, suggest products I should buy and even dangle discounts in front of my face.

That's a real problem, because the data they store on me is total bollocks.

Data retrieval
How big tech manages your personal information

Here's the real me: I'm a 33-year-old technology journalist who is married and owns his own home -- at least if you think having a 35-year mortgage qualifies as "ownership." I have one daughter who is now a few years old and a newborn son. I drive a third- or fourth-hand, petrol-powered Lexus from the year 2000, making my car old enough to vote.

Maybe it's egomania, but because I'm on the internet as much as I am and also because I have a semi-prominent job, I assumed these companies would know a lot about me. Like, I'm on Facebook and Twitter pretty much nonstop, and I spend 10 or 12 hours per day on the internet. How can I not be the most open book to these people?

And yet...

The Acxiom version of Dan Cooper has the same address as me, but he's 25, not 33, and he's cohabiting rather than married. Well, according to the documents, there's a 37.2 percent chance he's cohabiting and a 37.1 percent chance he's hitched. The data also shows that he's an "empty nester," meaning that his kids have grown up and moved out, which is just about plausible if he fathered his first child at seven.

Acxiom's data contradicts itself on a number of occasions, thanks to the weird, algorithmic way it was gathered. He's childless, but at the same time his kids were born in 1983 and 1986 -- clearly this alterna-Cooper blossomed early. He also somehow managed to buy a home in 1993, aged 11, which he mortgaged as a second-time buyer. I mean, I have to admit, this guy clearly has his life and priorities sorted out.

All of that time spent having kids while still in school and buying a house has, unfortunately, had a domino effect on his career. Acxiom's statistical models think that he's an employed individual (85.4 percent) working inside the socio-economic grade C2/D. That means he's either a skilled or unskilled manual laborer. The data also isn't clear about his politics: He either reads the reactionary, right-wing rag Daily Express or the center-left Guardian.

What the hell is a C2/D social grade?

Talk about the ossified nature of British society for long enough and the chat will always come back to the Class System. The country's immovable social strata means that upward mobility, based on some sense that hard work is rewarded, is limited. When Prince William married Catherine Middleton, the joke was that she was a "commoner" made good, despite being from a historically wealthy and powerful British dynasty.

Britain's obsession with class runs so deep as to inform the basis of the National Readership Survey (NRS) social grades. The survey was launched in 1956 as a way for the burgeoning ad industry to research the attitudes and wealth of its readers. The six NRS grades run from A through E, with C split in two to better define the top and bottom halves of the population. The categories are broken down by how you earn your money, with professionals marked as A, middle managers as B and so on.

If you work in politics or advertising or have any sort of interest in the nature of British society, then it's quite likely that you'll have heard the phrase "ABC1." That's a general shorthand for the upper and middle classes, while "C2DE" denotes the middle and lower orders. Ironically, despite being used in Acxiom's files, the company's former data chief, Dawn Orr, said that the NRS "simply isn't up to the job anymore."

Part of me is naturally delighted that this huge data brokerage has no clue who Dan Cooper really is. Through some combination of genius, trickery or luck, I have avoided becoming a cog in the Man's Machine. I've hidden in plain sight, and these companies don't know me, my lifestyle or my purchasing habits. Then again, the fact that Acxiom failed to get facts right that are a matter of public record is pretty concerning. Perhaps that's the reason why the company declined to participate in this story.

But then, I also wonder how far this information will travel and to what ends it will be used, especially in light of what's going on in other parts of the world. You may have heard of China's plans for a Social Credit System that combines big data and mass surveillance. It is, of course, a way to both monitor and tightly control the population, ensuring compliance with whatever policy is currently en vogue.

Already, China has blacklists that can prevent people from traveling, blocking job offers and making people's children ineligible to attend top schools. And there was even an attempt at letting the private sector run parts of the scheme: In 2015, Alibaba and Tencent were asked to test their own systems. Two years later, regulators realized that companies that have a profit motive might charge folks to bolster their score.

The point is that even now companies and governments are working out ways of using this sort of aggregated data for their ends. I'm reminded of the movie Brazil, in which a vast bureaucracy would rather assassinate its own citizens than rectify errors in its records. Scoff all you want, but it's entirely plausible that a government would simply poll ad-agency data, credit scores and social media information to begin profiling its citizens.

That leaves me at something of a crossroads: Do I kick up a fuss and demand greater accuracy in the surveillance? Or is it better to try, as best possible, to obfuscate the system in the hopes of making it unusable by any future government?

Data Retrieval credits
Features editor: Aaron Souppouris
Lead reporter: Chris Ip
Additional reporting: Matt Brian, Dan Cooper, Steve Dent, Jamie Rigg, Mat Smith, Nick Summers
Copy editor: Megan Giller
Illustration: Koren Shadmi