CHAPTER 1

The Digital YOU

TODAY, MANY MILLIONS of people like you and me generate at least 10different types of electronic data before our second cup of coffee. It startsfrom the moment we use a card key to gain access to our office and continues aswe listen to our voice mail, respond to that first e-mail, and so on. Most ofthese actions are just minutiae, about as memorable as where we ate lunch twoweeks ago. Nor are these actions widely visible the way they once were, backwhen office workers spent most of their days interacting directly rather thanhunched over their keyboards. Yet, paradoxically, the "digital breadcrumbs"these actions leave behind can accurately capture individual behavior in moredetail than ever before—not to mention permanently.

The trail doesn't stop when the workday is over. In fact, for some, that'sreally when the action starts. Once we're out of the office, we can tweet awayto our heart's content or spend hours commenting on all of our friends' Facebookupdates without fear of getting busted by the boss. Nevertheless, the secondparadox of the digital breadcrumb trail is the silent, often terrifying way itdemolishes the traditional boundaries between our personal and professionallives—while at the same time erecting even more impenetrable barriers.

Let me explain. Perhaps the best example of this second paradox in action is thejob seeker. In many corporations, résumés land in a hiring manager's inbox onlyafter an underling has found and attached the relevant Google, LinkedIn, andFacebook results for the applicant. All of this represents an amalgam of whatthe job seeker wishes potential employers to know about her, what the Internet"knows" about her, and what she allows her friends and family to know about her.The final effect is—well, unknowable. Imagine, for example, being filteredout of the set of top applicants for your dream job because you really likebowling or have a passion for collecting antique dolls. Or because a friend justcouldn't resist taking a really unflattering picture of you and postingit on the photo-sharing site Flickr. (If this sounds preposterous, remember thatwhen there are way too many generally qualified applicants for a given position,employers have to filter the list somehow.)

It is often said that the Internet is the most interactive medium ever. As such,if you ignore it, it will largely—if not totally—ignore you. Butconversely, the more time you actively spend interacting withit—depositing different kinds of breadcrumbs as you post content, respondto blog postings, join social networking sites, and so on—the larger yourprofile becomes over time. The result is that some people acquire pages andpages of Google results that are really about them (as opposed to someone elsewho happens to have the same name), while others seem to maintain totalanonymity. Thus, many of today's job seekers stumble on a new and insidious typeof stature gap, one that is purely digitally driven and seems a lot like abottomless pit if you are on the wrong side of it.

In the old days, you might buy a snazzy new suit and borrow your friend's muchnicer car to go to an important interview—and, of course, spend lots oftime trying to inject as much stature as possible into your résumé. But what isa coherent response to Google determining that you're obscure—or that forall intents and purposes, you don't exist at all? Is that a better fate thanbeing deemed uncool or unenlightened, a bit too much fun, or too much of aknow-it-all? If you spend lots of time sprucing up your LinkedIn or Face-bookpage so you'll have a larger digital footprint with more (hopefully good andconsistent) personality, but Google still thinks you don't really exist,do you end up looking even more pathetic and insignificant?

This dynamic gets scary pretty quickly.

The single most important thing to understand about the digital world is this:it is a place that is both enticing and dangerous in much the same way as is aforeign country in the physical world. Its seductions, such as its immenseconvenience, are much clearer than most of its dangers and petty mischiefs, evento the frequent traveler. Further, as we'll see, it is much safer to be atourist than a business traveler.

My perspective on all of this is not just that of an Internet user or even aSilicon Valley technologist. I run a company that is probably unlike anyother—an evidence analytics firm. When particularly large scandals,investigations, or lawsuits hit large companies, they hire us—not just forour patented software that analyzes the patterns in many millions of electronicdata records, but often for our staff of mathematicians, computationallinguists, fraud analysts, and other specialists who use this software todetermine who's been naughty or nice.

As a result, since 2002 we've looked at hundreds of thousands of interesting e-mailsand other types of electronic data and at analyses of millions more.Possibly even some of yours, dear reader! Because what few people realize isthat for every glamorous lawsuit or investigation that makes it above the foldin the Wall Street Journal, there are at least 10 times as many thatinvolve "normal" people—often without their knowledge.

For example, if you—or your company—are trying to sue one of ourclients, and we have a lot of your e-mails, instant messages (IMs), documents,phone calls, and other data to analyze, you have reason to worry. We speak morethan 20 different languages. Not only that, but our software is smart enough toknow that while, for example, you may always write in Russian to Grandma, yougenerally don't use that language in your professional life. So if you suddenlybegin, perhaps you are trying to subvert a compliance monitoring program orprying human eyes. The software will automatically detect everything fromcanceled meetings to missing reports to deviations from standard workflows. Itwill likewise reconstruct party invitation lists and determine which employeessocialize in their off-hours—as well as which ones used tosocialize but no longer do. We'll bust you for contradicting yourself, even intrivial ways.

And that's just in the first couple of days.

We also work with corporations to find any rotten apples that may be in theirbarrel. We work with investigators and some types of plaintiff s' firms.Sometimes we work with white-collar criminal defendants, ranging from technologyexecutives to Deborah Jeane Palfrey (a.k.a. the D.C. Madam), to try to helpestablish their innocence. Or at least to help them paint a more nuanced pictureof reality.

At Cataphora, we are paid to do what we like to call "corporate archaeology."This means we reconstruct often-complex real-world events on the basis ofsurviving data records—often millions of them. These comprise e-mailmessages, IMs, documents, phone logs, and many other kinds of data records, aswell as whatever relevant traces might still be had on the Internet (seeFigure 1.1). But in a deeper sense, what we really do is figure out whatmakes the target set of people tick. Does it seem probable that Margaret couldhave stolen a long look at the patent draft of a competitor? Would shehave? Did Terry lie to his investors, or was he more likely merely beingcareless or inept? Did Robert know that the discounted merchandise wasdefective, or was he probably far too drunk or hung over to have noticed? Oftenwe examine people's actions during times of great difficulty, when everythingseems to be unraveling around them. This tends to magnify personality traits,both good and bad, making them a bit easier to assess.

The work can be likened to reading different pieces of the cross-referenceddiaries of insanely prolific diarists. However, the overall effect is vastlymore potent, in part because so many real-life actions seem to be taken almoston autopilot. As such, a portrait constructed from these actions captures theessence of a person's nature with far greater clarity than anything he'sconsciously composed ever could. The fact that this portrait is painted withsophisticated computer programs adds the imprimatur of objectivity.

The attorneys and investigators with whom we work use our variouscomputer-generated archaeological exercises and personal portraits in theirwitness interviews, depositions, and even in trials. Key witnesses and actors ina big case expect to be asked certain obvious questions relevant to that case.Their attorneys will coach them accordingly. What they don't expect is some off-kilter question about why they changed some subtle habit at a particular pointin time. In other words, "When did you first start to suspect that there mightbe a defect in the product?" is obviously going to be a key question in aproduct liability case. But a question such as "Why did you stop having lunchwith James on Fridays?" or "Why did you start deleting e-mails at the end ofevery month?" often flummoxes opposing witnesses into a perplexed stupor thatbenefits our clients as they cross-examine them.

After we had been doing this work for a while, we began to understand that weknew more—really knew more—about our targets than theirspouses or closest friends did. Perhaps more than they knew about themselves. Wenot only knew whether they were happy or unhappy, but what made them that wayand how they behaved in either state. We could observe the consistency of theirviewpoints over time and whether they expressed the same opinions regardless ofwho was listening. We saw who was overgenerous about sharing credit and whoendeavored to grab credit for anything short of inventing the Internet. We sawwho was quick to lash out in anger and who was ready to forgive and forget, whowas sincere and who was spiteful. We saw who did work and who mostly justcomplained about the work done by others. Most of all though, we saw whocommanded respect, influence, and loyalty—as well as who was in no realdanger of ever commanding any of these. And who had the self-awareness to knowwhere he or she truly stood in the esteem of others.

Eventually, we came to realize something more: we were clinically capturing thecharacter of both individuals and organizations. Really understanding thecharacter of the key players is essential in many types of litigation,specifically those in which the facts themselves are unknown or in dispute. Ofcourse there are myriad other reasons why such information is interesting toemployers, banks and other institutions that might extend you credit, insurancecompanies, marketers, people you are dating, those with whom you might dobusiness, and so on. Our job is made easier in the litigation context, becausethe corporations involved provide us with all of the electronic data for therelevant people, but much can be learned just on the basis of what's publiclyavailable. The more ways that people feel almost compelled to participate on theInternet, the greater the prospects for analysis.

To go back to our résumé example, a website called Emurse urges you to use itssite to ensure that only one version of your résumé is accessible on theInternet; this is designed to prevent various types of possible problems andembarrassments. For example, if different versions of your résumé are floatingaround out there in cyberspace, the fact that you took the liberty of slightlyembellishing your title in a prior job—perhaps promoting yourself toproject lead—is likelier to be noticed. Even by a person. A computerprogram can do this easily; it recognizes you by a combination of your name andother key information, such as details of your college degrees. Identifying aWeb page or document as a résumé is also an easy trick for software, sincerésumés almost always contain certain common types of phrases and content, suchas "education" or "work history." Detecting inconsistencies in the dates ofemployment or the phrasing of a title associated with a particular job indifferent versions of the résumé is only slightly more difficult; the same istrue for catching verb "upgrades," such as "participated in a team of 20" versus"led a team of 20."

Extending this example a bit further, imagine a résumé that has been upgradedseveral times over a period of months. The perpetrator now looks like adesperate liar and will likely be left with both the time and the need toupgrade quite a few more times as a result. Likewise, several versions of arésumé that have each been tailored to appeal to a specific employer or type ofemployer can easily seem obsequious and insincere in this context, even thoughsuch tailoring is often encouraged by career coaches.

This is just one narrow example, however. Many others spring tomind—evolving profiles on dating sites, finance-related newsgroups, andother forums that over time can expose you as having been cosmically wrong, suchas when you confidently declared that it was obvious that a particularpresidential candidate would win the general election in a landslide only tohave the candidate wipe out early in the primaries.

Tack all of the personal digital breadcrumb–generating activity on to theliterally billions of e-mails, IMs, and other forms of electronic communicationthat occur every day during the workweek, and you have millions of digitalportraits capturing every aspect of people's lives. Any current employer whowishes to do so can easily amass much or all of this information; in the UnitedStates, any data you create on your employer's dime and time belongs to them. Inother words, privacy really doesn't apply when you are using your employer'scomputers and network. As the percentage of the workday spent online continuesto grow, more and more employers will find themselves in the position of needingto take further steps to monitor their employees. The motivations are myriad:preventing fraud, reducing some kinds of insurance premiums, determining who isproductive, or simply better understanding the actual functioning of theirbusiness.

What can be known about you from your digital bread-crumb trail, and how? Let'sstart off with a simple example. The data visualization in Figure 1.2,which is from the free Digital Mirror software available atdigitalmirrorsoftware.com, illustrates several possible ways ofassessing the importance that different individuals assign to one another basednot on any company organizational chart or social network links proclaimingeternal friendship, but solely on empirical assessment of behavior. The purposeis to identify relative differences between how one person treats theindividuals he interacts with the most and how these same people treat him.

Some of these categories require merely tallying data records (like counting upthe number of IMs or phone calls exchanged), while others require theapplication of computational linguistics techniques. For example, accuratelydetecting an "I'm too busy" response requires automatically determining thedifference between someone saying, "Things are going great, but I'm reallybusy," and "I don't have time to go out to lunch with you because I'm too busy."Determining when someone is soliciting an opinion is also a bit involved; we'llcome back to this in a later chapter. For now though, the main point to note isthat we try never to draw a conclusion based solely on any one kind ofmeasurement. Reality is complicated; multiple probes—and multiple types ofprobes—of the same question are necessary to get a reasonably correctanswer for most things.

This type of coldhearted, objective analysis underscores the fact that none ofus treats all of the actors in our professional or personal lives equally. Noteven close, in fact. We eagerly grab the chance to interact with some, whilestriving to duck out of others' sight lines—just as other people may do tous. Still, when people see this kind of analysis done on themselves, a commonreaction is that they can't believe they were so consistently rude to this orthat colleague. "Surely there's some mistake," they usually say.

No, there isn't.

It is simply a matter of not having previously assessed reality in these terms.Most people tend to think in the old-fashioned way, at the unit level of theindividual phone call or e-mail. They may know they did not return a particularphone call as quickly as they might have. But they don't think of it inquantitative terms; for example, how much faster they typically respond to Suzythan to Joan. Likewise, while you may realize that you tend to brush off anoverly talkative coworker when you are especially pressed, you may not realizethat you nearly always do it. Or that you sometimes leave behind a mile-wide digital breadcrumb trail on a corporate wiki—or anywhere else thetalker may be able to see it—just after you've told him or her you areheaded into a meeting for the rest of the day.

So what is the Digital YOU then? It is much more than the sum of the breadcrumbsyou leave behind in your travels through cyberspace. The Digital YOU is acomplex mosaic of habit, subconscious acts of both omission and commission, andpremeditated presentations. It is how your peers and friends, bosses and familyactually experience you, as more and more of your life in the real world takesplace in the digital one. The purpose of this book and the associated softwarethat you can download from our website is to help you begin to experience yourDigital YOU in the same way everyone around you does.

* * *

A final note: We work under the umbrella of attorney-client privilege,just like the lawyers on TV. So we can only name names when our clients give uspermission to do so, and most cannot, because the matters in question are stillpending as of this writing. However, all of the various anecdotes included inthis book come from either a case in which we were involved or a case that oneof the law firms with which we regularly work contributed. In a few cases, weused something we read about in the newspaper that was too good not to include.

The software industry stories all come from the direct experience of someone atCataphora or one of our friends.

(Continues...)