February 21, 2012

Four years ago I saluted the New York Times for giving front page coverage to the rape of consumers' personal data and declared "consumers can no longer ignore the extent of this phenomenon". Today few would deny it.

This awareness is partly fueled by the tenfold increase of Facebook's valuation in the same period. As Lori Andrews writes, "Facebook's inventory consists of [nothing but] personal data - yours" (*). When it sells to advertisers, "Facebook is using you", a fact hard to hide to "845 million users".

And yet people must have remained naive at heart. If not, why would Nicole Perlroth and Nick Bilton take pains to pen an article on the fact that "the address book in smartphones [...] is free for app developers to take at will, often without the phone owner's knowledge" (**). The authors have shared a dozen of names of such self-services with their readers. If you are shocked, you have not been reading these fillips for very long.

Still, for many users, this is not a big deal and personal data pirates do count on such complicit acquiescence. When caught, they are wont to say users ask for it. Notice though how pirates hardly trust in the truth of what they say. If their victims were so willing to part with their data, why do they resort to the usual threat, no data, no service, even when their service does not require it?

Whenever possible, corporate pirates prefer to let accomplices do the dirty job. Read Richard Falkenrath about his children's school (***). "Like a growing number nationwide, [it] bought its new email system with the currency of its students' future, [...] the long-term monetisation of [their] data".

Corporate pirates' plunder represent sizable stakes. Even more profitable is squeezing serfs forbidden to leave the land they till for their masters.

Or else why "big tech companies" would be so eager "to be all things to all people all day long", as aptly put by David Streitfeld (****)? Remember the walled gardens created to keep users enthralled? It turns out the keyword here is "wall" not "garden". "It is probably not a coincidence that much of the lingo of these systems borrows from prison, starting with "locked in"".

The benefits accruing to the would-be jailors are highlighted in Steve Lohr's essay on "the Age of Big Data" (*****). Large companies "adopting "data-driven decision making" achieved productivity gains that were 5 percent to 6 percent higher than other factors could explain". It does not seem much but scale is everything. Applied to economic output, the same lowly percentage gain separates stagnant economies from healthy ones.

The long term risks to captive consumers are proportional. Echoing Joseph Turow's concerns, Lori Andrews warns us "stereotyping is alive and well in data aggregation". To misclassification, add the danger of mismodeling. As Steve Lohr reminds us, "Big Data has its perils, to be sure. With huge data sets and fine grained measurement, [...] there is increased risk of "false discoveries"". Error is inherent to statistics.

"Despite the caveats, there seems to be no turning back. Data is in the driver's seat." This after all is the Information Age.

Should we regret it? That progress brings costs as well as benefits is not reason enough to condemn it. But notice Steve Lohr states that the driver is "data", not "Big Data". The question is whether progress requires society to keep the costs while Big Data reap the profits.

Big Data needs not present itself as flawless. It has only to claim to be inevitable. If it is not so as far as technology is concerned, it could still be so from a financial perspective. "Consumer information is worth billions in aggregate, but individually, the bits of data are worth practically nothing" is how Joshua Brustein neatly encapsulates the issue (******).

It is not for want of trying. Joshua Brustein lists five start-ups whose business model implies personal data should benefit the person it describes. I approve. Indeed I claim most of us are called to manage our personal information for profit. Eprivacy is not a luxury, it ought to be a source of jobs.

Not that it will be easy as "companies don't need to pay for the information when they get it for free". It is the polite way to speak of corporate digital piracy and this practice will remain legal as long as laws are for sale in our western pronaocracies.

But the very nature of statistics creates a second obstacle. No matter how well informed, a supplier runs a risk when it spends money to contact a consumer. Hiding each gamble in its statistical average, aggregation guarantees success to well managed marketing campaigns. Nor can pattern recognition models be computed from isolated samples while they can be from their aggregation.

One response is to organize consumers. Unions and labor laws were thus set up to protect individual workers from dominant employers. But perhaps the most promising avenue is to realize consumers do not act in isolation even today. Call it word of mouth or recommendations, most marketing expenses go into modifying existing recommendation networks, "the way we live now", and creating new ones, "for a better tomorrow".

Recommendations are instances of the multiple links which bind together the members of a given society and represent an intangible economic value some call social capital (1). If individuals were able to monetize such contributions at their real value, wouldn't marketing productivity rise just as industrial productivity did in the Energy Age, and, with it, wouldn't new social benefits grow for ever, no less tangible than ordinary salaries?

That greater efficiency can arise from a partial monetization of social capital is anathema to people like Yochai Benkler. For them money corrupts relationships by warping individual behaviors from selfless contributions towards selfish gains. Though well intentioned, their analysis is faulty. For it is naive to think individuals act as selfless saints as long as money is absent. Saints are honored precisely for being outside the norm.

Yochai Benkler is right in a limited sense only. If the recipient of a service feels no longer obligated to the giver or comes to suspect his intention, monetization does risk destroying social capital by loosening the links between members. We must be conscious of the danger but we should not overblow it. While no doubt some teachers are mercenary, who wants to go back to a time when teachers received no money for their pains?

Done right, monetization of distributed social capital can dispense the benefits of data aggregation while dispensing with the risks of Big Data.

Philippe Coueignoux

• (*) ........... Facebook Is Using You, by Lori Andrews (New York Times) - Feb 5, 2012
• (**) ......... An Easy Sweep Of User Data From Phones, by Nicole Perlroth and Nick Bilton (New York Times) - Feb 16, 2012
• (***) ....... Google must remember our right to be forgotten, by Richard Falkenrath (Financial Times) - Feb 16, 2012
• (****) ..... Seeking the Captive Customer, by David Streitfeld (New York Times) - Feb 13, 2012
• (*****) ... The Age of Big Data, by Steve Lohr (New York Times) - Feb 12, 2012
• (******) . Start-Ups Aim to Help Users Put a Price on Their Personal Data, by Joshua Brustein (New York Times) - Feb 13, 2012
• (1) see social capital in the wikipedia