I’m Being Followed: How Google — and 104 Other Companies — Track Me on the Web

Alexis Madrigal, The Atlantic
Add to Briefcase
Alexis Madrigal, The Atlantic
March 1, 2012, 6:58 a.m.

On Thursday morn­ing, if you opened your browser and went to, an amaz­ing thing happened in the mil­li­seconds between your click and when the news about North Korea and James Mur­doch ap­peared on your screen. Data from this single vis­it was sent to 10 dif­fer­ent com­pan­ies, in­clud­ing Mi­crosoft and Google sub­si­di­ar­ies, a gaggle of traffic-log­ging sites, and oth­er, smal­ler ad firms. Nearly in­stant­an­eously, these com­pan­ies can log your vis­it, place ads tailored for your eyes spe­cific­ally, and add to the ever-grow­ing on­line file about you.

Latest Polit­ics Posts:
Load­ing feed…

There’s noth­ing ne­ces­sar­ily sin­is­ter about this sub­ter­ranean data ex­change: This is, after all, the ad­vert­ising eco­sys­tem that sup­ports free on­line con­tent. All the data lets ad­vert­isers tune their ads, and the rest of the in­form­a­tion log­ging lets them meas­ure how well things are ac­tu­ally work­ing. And I do not mean to pick on The New York Times. While vis­it­ing The Huff­ing­ton Post or The At­lantic or Busi­ness In­sider, the same pro­cess hap­pens to a great­er or less­er de­gree. Every move you make on the In­ter­net is worth some tiny amount to someone, and a panoply of com­pan­ies want to make sure that no step along your In­ter­net jour­ney goes un­mon­et­ized.

Even if you’re gen­er­ally fa­mil­i­ar with the idea of data col­lec­tion for tar­geted ad­vert­ising, the num­ber and vari­ety of these data col­lect­ors will prob­ably as­ton­ish you. Al­low me to in­tro­duce the list of com­pan­ies that tracked my move­ments on the In­ter­net in one re­cent 36-hour peri­od of stand­ard Web surf­ing: Acerno. Adara Me­dia. Ad­blade. Ad­brite. ADC Onion. Ad­chemy. ADi­FY. Ad­Meld. Adtech. Ag­greg­ate Know­ledge. Al­mond­Net. Aper­ture. AppNex­us. At­las. Audi­ence Sci­ence.

And that’s just the A’s. My com­plete list in­cludes 105 com­pan­ies, and there are dozens more than that in ex­ist­ence. You, too, could com­pile your own list us­ing Moz­illa’s tool, Col­lu­sion, which re­cords the com­pan­ies that are cap­tur­ing data about you, or more pre­cisely, your di­git­al self.

While the big names — Google, Mi­crosoft, Face­book, Ya­hoo, etc. — show up in this cata­log, the bulk of it is com­posed of smal­ler data and ad­vert­ising busi­nesses that form a shad­ow web of com­pan­ies that want to help show you ad­vert­ising that you’re more likely to click on and products that you’re more likely to pur­chase.

To be clear, these com­pan­ies gath­er data without at­tach­ing it to your name; they use that data to show you ads you’re stat­ist­ic­ally more likely to click. That’s the game, and there is sub­stan­tial money in it.

As users, we move through our In­ter­net ex­per­i­ences un­aware of the churn­ing sub­ter­ranean ma­chines power­ing our Web pages with their cook­ies and pixels track­ers, their track­ing code and data­bases. We shop for wed­ding cater­ers and sud­denly see ring ads ap­pear on ran­dom Web pages we’re vis­it­ing. We some­times think the ads fol­low­ing us around the In­ter­net are “creepy.” We some­times feel watched. Does it mat­ter? We don’t really know what to think.

The is­sues the in­dustry raises did not ex­ist when Ron­ald Re­agan was pres­id­ent and were only in nas­cent form when the Twin Towers fell. These are phe­nom­ena of our time, and while there are many ante­cedent forms of ad­vert­ising, nev­er be­fore in the his­tory of hu­man ex­ist­ence has so much data been gathered about so many people for the sole pur­pose of selling them ads.

“The best minds of my gen­er­a­tion are think­ing about how to make people click ads,” my old friend and an early Face­book em­ploy­ee Jeff Ham­merbach­er once said. “That sucks,” he ad­ded. But in­creas­ingly I think these is­sues — how we move “freely” on­line, or more prop­erly, how we pay one way or an­oth­er — are ac­tu­ally the lead­ing edge of a much big­ger dis­cus­sion about the re­la­tion­ship between our di­git­al and phys­ic­al selves. I don’t mean the­or­et­ic­ally or psy­cho­lo­gic­ally. I mean that the norms es­tab­lished to im­prove how of­ten people click ads may end up de­term­in­ing who you are when viewed by a bank or a ro­mantic part­ner or a re­tail­er who sells shoes.

Already, the web­sites you vis­it re­shape them­selves be­fore you like a car­ni­vor­ous school of fish, and this is only the be­gin­ning. Right now, a huge chunk of what you’ve ever looked at on the In­ter­net is sit­ting in data­bases all across the world. The line sep­ar­at­ing all that it might say about you, good or bad, is as thin as the let­ters of your name. If and when that wall breaks down, the num­bers may over­whelm the name. The un­con­sciously cre­ated pro­file may mean more than the ex­amined self I’ve sought to build.

Most pri­vacy de­bates have been couched in the tech­nic­al. We read about how Google by­passed Sa­fari’s pri­vacy set­tings, whatever those were. Or we read the de­tails about how Face­book tracks you with those friendly “Like” but­tons. Be­hind the de­tails, however, are a tangle of philo­soph­ic­al is­sues that are at the heart of the struggle between pri­vacy ad­voc­ates and on­line ad­vert­ising com­pan­ies: What is an­onym­ity? What is iden­tity? How sim­il­ar are hu­mans and ma­chines? This es­say is an at­tempt to think through those ques­tions.

The bad news is that people haven’t taken con­trol of the data that’s be­ing col­lec­ted and traded about them. The good news is that — in a quite lit­er­al sense — simply think­ing dif­fer­ently about this ad­vert­ising busi­ness can change the way that it works. After all, if you take these com­pan­ies at their word, they ex­ist to serve users as much as to serve their cli­ents.


Be­fore we get too deep, let’s talk about the real­ity of the on­line dis­play ad­vert­ising in­dustry. (That means, es­sen­tially, all the ads not as­so­ci­ated with a Web search.) There are a dizzy­ing ar­ray of com­pan­ies and ser­vices who can all make a buck by help­ing ad­vert­isers tar­get you a teensy, weensy bit bet­ter than the next guy. These are com­pan­ies that must prove them­selves quite nar­rowly in meas­ur­able rev­en­ue and profit; the com­pet­i­tion is fierce, the prize is large, and the strategies are ever-chan­ging. Here’s the cor­al-reef level di­versity of cor­por­ate life in dis­play ad­vert­ising, as cata­loged by Luma Part­ners a little over a year ago:

Don’t get too caught up in all of that, though. There are three ba­sic cat­egor­ies: Es­sen­tially, there are people who help the buy­ers (on the left), people who help the sellers (on the right), and a whole lot of people who as­sist either side with more data or faster ser­vice or bet­ter meas­ure­ment. Let’s zoom in on three of them — just from the A’s — to give you an idea of the kinds of out­fits we’re talk­ing about.

Let’s look at three com­pan­ies from our list of A’s. Ad­netik is a stand­ard tar­get­ing com­pany that uses real-time bid­ding. They can of­fer tar­geted ads based on how users act (be­ha­vi­or­al), who they are (demo­graph­ic), where they live (geo­graph­ic), and who they seem like on­line (look-alike), as well as something they call “so­cial prox­im­ity.” They also give ad­vert­isers the abil­ity to choose the types of sites on which their ads will run based on “para­met­ers like pub­lish­er brand equity, con­tex­tu­al rel­ev­ance to the ad­vert­iser, brand safety, level of ad clut­ter, and con­tent qual­ity.”

It’s worth not­ing how dif­fer­ent this prac­tice is from tra­di­tion­al ad­vert­ising. The so­cial con­tract between ad­vert­isers and pub­lic­a­tions used to be that pub­lic­a­tions gathered par­tic­u­lar types of people in­to something called an audi­ence, then ad­vert­isers pur­chased ads in that pub­lic­a­tion to reach that audi­ence. There was an art to it, and some pub­lic­a­tions had cachet while oth­ers didn’t. On­line ad­vert­ising upends all that: Now you can buy the audi­ence without the pub­lic­a­tion. You want an At­lantic read­er? Great! Some ad net­work can sell you someone who has been to The At­lantic but is now read­ing about hand lo­tion at Know­YourHand­Lo­ And they’ll sell you that set of eye­balls for a fifth of the price. You can bid in real time on a set of those eye­balls across mil­lions of sites without ever talk­ing to an ad­vert­ising sales­per­son. (Of course, such a tradeoff has costs, which we’ll see soon.)

Ad­netik also of­fers a ser­vice called “re­tar­get­ing” that an­oth­er A com­pany, Ad­Roll, spe­cial­izes in. Here’s how it works. Let’s say you’re an on­line shoe mer­chant. Someone comes to your store but doesn’t pur­chase any­thing. While they’re there, you drop a cook­ie on them. There­after you can tar­get ads to them, know­ing that they’re at least mildly in­ter­ested. Even bet­ter, you can drop cook­ies on every­one who comes to look at shoes and then watch to see who comes back to buy. Those people be­come your train­ing data, and soon you’re only “re­tar­get­ing” those people with a data pro­file that in­dic­ates that they’re likely to pur­chase something from you even­tu­ally. It’s slick, es­pe­cially if people don’t no­tice that the pairs of shoes they found the will­power not to pur­chase just hap­pen to be show­ing up on their fa­vor­ite garden­ing sites.

There are many power­ful things you can do once you’ve got data on a user, so the big wor­ries for on­line ad­vert­isers shift to the in­vent­ory it­self. Pur­chas­ing a page in a magazine is a pro­cess through which ad­vert­isers have sig­ni­fic­ant con­trol; but these types of on­line ads could con­ceiv­ably run any­where. After all, many ad net­works need all the in­vent­ory they can get, so they sign up all kinds of con­tent pro­viders. And that’s where our third com­pany comes in­to play.

Ad­Ex­pose, now a com­Score com­pany, watches where and how ads are run to de­term­ine if their pur­chasers got their money’s worth. “Up to 80 per­cent of in­ter­act­ive ads are sold and resold through third parties,” they put it on their web­site. “This daisy­chain­ing brings down the value of on­line ads and ad­vert­isers don’t al­ways know where their ads have run.” To solve that prob­lem, Ad­Ex­pose claims to provide in­de­pend­ent veri­fic­a­tion of an ad’s place­ment.

All three com­pan­ies want to know as much about me and what’s on my screen as they pos­sibly can, al­though they have dif­fer­ent reas­ons for their in­terest. None of them seem like evil com­pan­ies, nor are they sin­gu­lar com­pan­ies. Like much of this in­dustry, they seem to be­lieve in what they’re do­ing. They de­liv­er more rel­ev­ant ad­vert­ising to con­sumers and that makes more money for com­pan­ies. They are simply tools to im­prove the grip strength of the in­vis­ible hand.


And yet, the rev­el­a­tion that 105 dif­fer­ent out­fits were col­lect­ing and pre­sum­ably selling data about me on the In­ter­net gives me pause. It’s not just Google or Face­book or Ya­hoo. There are lit­er­ally dozens and dozens of these com­pan­ies and the av­er­age user has no idea what they do or how they work. We just know that for some reas­on, at one point or an­oth­er, an or­gan­iz­a­tion dropped a cook­ie on us and have cre­ated a file on some serv­er, stead­ily ac­cu­mu­lat­ing clicks and habits that will even­tu­ally be mined and mar­keted.

The on­line ad­vert­ising in­dustry ar­gues that tech­no­logy is chan­ging so rap­idly that reg­u­la­tion is not the an­swer to my queas­i­ness about all that data go­ing off to who-knows-where. The prob­lem, however, is that the in­dustry’s ver­sion of self-reg­u­la­tion is not one that most people would ex­pect or agree with, as I found out my­self.

After run­ning Col­lu­sion for a few days, I wanted to see if there was an easy meth­od to stop data col­lec­tion. Na­ively, I went to the self-reg­u­lat­ory site run by the Net­work Ad­vert­ising Ini­ti­at­ive and com­pleted their “Opt Out” form. I did so for the dozens of com­pan­ies lis­ted and I would say that it was a simple and nom­in­ally ef­fect­ive pro­cess. That said, I wasn’t sure if data would stop be­ing col­lec­ted on me or not. The site it­self does not say that data col­lec­tion will stop, but it’s also not clear that data col­lec­tion will con­tin­ue. In fact, the over­view of NAI’s prin­ciples freely mixes talk about how the or­gan­iz­a­tion’s code “lim­its the types of data that mem­ber com­pan­ies can use” with in­form­a­tion about the opt-out pro­cess.

After opt­ing out, I went back to Col­lu­sion to see if com­pan­ies were still track­ing me. I found that many, many com­pan­ies ap­peared to be log­ging data for me. Ac­cord­ing to Moz­illa, the cur­rent ver­sion of Col­lu­sion does not al­low me to see pre­cisely what com­pan­ies are still track­ing, but Stan­ford re­search­ers us­ing Col­lu­sion found that at least some com­pan­ies con­tin­ue to col­lect data. All that I had “op­ted out” of was re­ceiv­ing tar­geted ads, not data col­lec­tion. There is no way, through the com­pan­ies’ own self-reg­u­lat­ory ap­par­at­us, to stop be­ing tracked on­line. None.

After those Stan­ford re­search­ers pos­ted their res­ults to a uni­versity blog, they re­ceived a sharp re­sponse from the NAI’s then-chief, Chuck Cur­ran.

In es­sence, Cur­ran ar­gued that users do not have the right to not be tracked. “We’ve long re­cog­nized that con­sumers should be provided a choice about wheth­er data about their likely in­terests can be used to make their ads more rel­ev­ant,” he wrote. “But the NAI code also re­cog­nizes that com­pan­ies some­times need to con­tin­ue to col­lect data for op­er­a­tion­al reas­ons that are sep­ar­ate from ad tar­get­ing based on a user’s on­line be­ha­vi­or.”

Com­pan­ies “need to con­tin­ue to col­lect data,” but that con­trasts dir­ectly with users de­sire “not to be tracked.” The only right that on­line ad­vert­isers are will­ing to give users is the abil­ity not to have ads served to them based on their web his­tor­ies. Cur­ran him­self ad­mits this: “There is a vi­tal dis­tinc­tion between lim­it­ing the use of on­line data for ad tar­get­ing, and ban­ning data col­lec­tion out­right.”

But based on the scant sur­vey and an­ec­dot­al data that we have avail­able, when users opt out, pre­vent­ing data col­lec­tion is pre­cisely what they are after.

In pre­lim­in­ary res­ults from a sur­vey con­duc­ted last year, Alee­cia Mc­Don­ald, a fel­low at Stan­ford Cen­ter for In­ter­net and So­ci­ety, found that users ex­pec­ted a lot more from the cur­rent set of tools than those tools de­liv­er. The largest per­cent­age of her sur­vey group (34 per­cent) who looked at the NAI’s opt-out page thought that it was “a web­site that lets you tell com­pan­ies not to col­lect data about you.” For browser-based “Do Not Track” tools, a full 61 per­cent of re­spond­ents ex­pec­ted that if they clicked such a but­ton, no data would be col­lec­ted about them.

Do Not Track tools have be­come a ma­jor point of con­ten­tion. The idea is that if you en­able one in your browser, when you ar­rive at The New York Times, you send a her­ald out ahead of you that says, “Do not col­lect data about me.” Mem­bers of the NAI have agreed, in prin­ciple, to fol­low the “Do Not Track” pro­vi­sions, but now the de­bate has shif­ted to the de­tails.

There is a fas­cin­at­ing scrum over what DNT tools should do and what or­ders web­sites will have to re­spect from users. The Di­git­al Ad­vert­ising Al­li­ance (of which the NAI is a part), the Fed­er­al Trade Com­mis­sion, W3C, the In­ter­net Ad­vert­ising Bur­eau (also part of the DAA), and pri­vacy re­search­ers at aca­dem­ic in­sti­tu­tions are all in­volved. In Novem­ber, the DAA put out a new set of prin­ciples that con­tain some good ideas like the pro­hib­i­tion of “col­lec­tion, use or trans­fer of In­ter­net surf­ing data across Web­sites for de­term­in­a­tion of a con­sumer’s eli­gib­il­ity for em­ploy­ment, cred­it stand­ing, health­care treat­ment and in­sur­ance.”

This week, the White House seemed to side with pri­vacy ad­voc­ates who want to lim­it col­lec­tion, not just uses. Its Con­sumer Pri­vacy Bill of Rights pushes com­pan­ies to al­low users to “ex­er­cise con­trol over what per­son­al data com­pan­ies col­lect from them and how they use it.” The DAA her­al­ded its own par­ti­cip­a­tion in the White House pro­cess, though even it noted this is the be­gin­ning of a long jour­ney.

There has been a clear and real philo­soph­ic­al dif­fer­ence between the ad­vert­isers and reg­u­lat­ors rep­res­ent­ing Web users. On the one hand, as Stan­ford pri­vacy re­search­er Jonath­an May­er put it, “Many stake­hold­ers on on­line pri­vacy, in­clud­ing U.S. and E.U. reg­u­lat­ors, have re­peatedly em­phas­ized that ef­fect­ive con­sumer con­trol ne­ces­sit­ates re­stric­tions on the col­lec­tion of in­form­a­tion, not just pro­hib­i­tions on spe­cif­ic uses of in­form­a­tion.” But ad­vert­isers want to keep col­lect­ing as much data as they can as long as they prom­ise to not to use it to tar­get ad­vert­ising. That’s why the NAI opt-out pro­gram works like it does.

Let’s not linger too long on the tech­nic­al im­ple­ment­a­tion here: There may be some top­ics around which com­prom­ises can be found. Some defin­i­tion of “Do Not Track” that suits in­dustry and pri­vacy people may be craf­ted. Vari­ous is­sues re­lated to dif­fer­ences between first- and third-party cook­ies may be re­solved. But the battle over data col­lec­tion and ad tar­get­ing goes much deep­er than the tac­tic­al, tech­nic­al is­sues that dom­in­ate the dis­cus­sion

Let’s as­sume good faith on be­half of ad­vert­ising com­pan­ies and con­front the core is­sue head-on: Should users be able to stop data col­lec­tion, even if com­pan­ies aren’t do­ing any­thing “bad” with it? Should that be a right, as the White House con­tends, and more im­port­antly, why?


Com­pan­ies’ abil­ity to track people on­line has sig­ni­fic­antly out­paced the cul­tur­al norms and ex­pect­a­tions of pri­vacy. This is not be­cause on­line com­pan­ies are worse than their off­line coun­ter­parts, but rather be­cause what they can do is so, so dif­fer­ent. We don’t have a lan­guage for talk­ing about how these com­pan­ies func­tion or how our so­ci­ety should deal with them.

The word you hear over and over and over is that tar­geted ads can be “creepy.” It even crops up in the aca­dem­ic lit­er­at­ure, des­pite its vague mean­ing in this con­text. My in­tu­ition is that we use the word “creepy” pre­cisely be­cause it is an in­de­term­in­ate word. It con­notes that tingling-back-of-the-neck feel­ing, but not ne­ces­sar­ily more than that. The creepy feel­ing is a sign to pay at­ten­tion to a pos­sibly harm­ful phe­nomen­on. But we can’t sort our feel­ings in­to cat­egor­ies — dan­ger­ous or harm­less — be­cause we don’t ac­tu­ally know what’s go­ing to hap­pen with all the data that’s be­ing col­lec­ted.

Not only are there more than 100 com­pan­ies that are col­lect­ing data on us, mak­ing it prac­tic­ally im­possible to sort good from bad, but there are key un­re­solved is­sues about how we re­late to our di­git­al selves and the ma­chines through which they are ex­pressed.

At the heart of the prob­lem is that we in­creas­ingly live two lives: a phys­ic­al one in which your name, So­cial Se­cur­ity num­ber, pass­port num­ber, and driver’s li­cense are your main iden­tity mark­ers, and one di­git­al, in which you have dozens of iden­tity mark­ers, which are known to you and me as cook­ies. These mark­ers al­low data gather­ers to keep tabs on you without your name. Those cook­ie num­bers, which are known only to the en­tit­ies that as­signed them to you, are per­sist­ent mark­ers of who you are, but they re­main un­at­tached to your phys­ic­al iden­tity through your name. There is a (thin) wall between the self that buys health in­sur­ance and the self that searches for health-re­lated in­form­a­tion on­line.

For real-time ad­vert­ising bid­ding, in which audi­ences are be­ing served ads that were pur­chased mil­li­seconds after users ar­rive at a Web page, ad ser­vices “match cook­ies,” so that both sides know who a user is. While that in­form­a­tion may not be stored by both com­pan­ies, i.e. it’s not ad­ded to a user’s per­sist­ent file, it means that the walls between on­line data selves are fall­ing away quickly. Every­one can know who you are, even if they call you by a dif­fer­ent num­ber.

Fur­ther­more, many com­pan­ies are just out there col­lect­ing data to sell to oth­er com­pan­ies. Any­one can com­bine mul­tiple data­bases to­geth­er in­to a fully fleshed out di­git­al por­trait. As a Wall Street Journ­al in­vest­ig­a­tion put it, data com­pan­ies are “trans­form­ing the In­ter­net in­to a place where people are be­com­ing an­onym­ous in name only.” Joe Tur­ow, who re­cently pub­lished a book on on­line pri­vacy, had even stronger words:

If a com­pany can fol­low your be­ha­vi­or in the di­git­al en­vir­on­ment — an en­vir­on­ment that po­ten­tially in­cludes your mo­bile phone and tele­vi­sion set — its claim that you are “an­onym­ous” is mean­ing­less. That is par­tic­u­larly true when firms in­ter­mit­tently add off-line in­form­a­tion such as shop­ping pat­terns and the value of your house to their on­line data and then simply strip the name and ad­dress to make it “an­onym­ous.” It mat­ters little if your name is John Smith, Yesh Mis­par, or 3211466. The per­sist­ence of in­form­a­tion about you will lead firms to act based on what they know, share, and care about you, wheth­er you know it is hap­pen­ing or not.

Mil­it­at­ing against this col­lapse of pri­vacy is a pro­tec­tion em­bed­ded in the very nature of the on­line ad­vert­ising sys­tem. No per­son could ever ac­tu­ally look over the world’s Web tracks. It would be too ex­pens­ive and even if you had all the hu­man laborers in the world, they couldn’t do the math fast enough to con­stantly re­cal­cu­late Web surfers’ value to ad­vert­isers. So, ma­chines are the ones that do all of the work.

When new tech­no­lo­gies come up against our ex­pect­a­tions of pri­vacy, I think it’s help­ful to make a real-world ana­logy. But we just do not have an ad­equate un­der­stand­ing of an­onym­ity in a world where ma­chines can parse all of our be­ha­vi­or without hu­man over­sight. Most ob­vi­ously, with the ma­chine, you have more pri­vacy than if a per­son were watch­ing your click­streams, pick­ing up col­lat­er­al know­ledge. A hu­man could eas­ily ap­ply ana­lyt­ic­al reas­on­ing skills to fig­ure out who you were. And any hu­man could use this data for un­au­thor­ized pur­poses. With our data-driv­en ad­vert­ising world, we are re­ly­ing on ma­chines’ cur­rent dumb­ness and in­ab­il­ity to “know too much.”

This is a double-edged sword. The cur­rent levels of ma­chine in­tel­li­gence in­su­late us from pri­vacy cata­strophe, so we let data be col­lec­ted about us. But we know that this data is not go­ing away and yet ma­chine in­tel­li­gence is grow­ing rap­idly. The res­ults of this pro­cess are in­eluct­able. Left to their own devices, ad track­ing firms will even­tu­ally be able to con­nect your vari­ous data selves. And then they will break down the name wall, if they are al­lowed to.


Your vis­it to this story prob­ably gen­er­ated data for 13 com­pan­ies through our web­site. The great down­side to this beau­ti­ful, free Web that we have is that you have to sell your di­git­al self in or­der to ac­cess it. If you’d like to stop data col­lec­tion, take a look at Do Not Track Plus. It goes bey­ond Col­lu­sion and browser based con­trols in block­ing data col­lec­tion out­right.

But I am ul­ti­mately un­clear what I think about us­ing these tools. Rhet­or­ic­ally, they im­ply that there will be tech­no­lo­gic­al solu­tions to these data col­lec­tion prob­lems. Un­doubtedly, tech elites will use them. The prob­lem is the vast ma­jor­ity of In­ter­net users will nev­er know what’s churn­ing be­neath their browsers. And the ad­vert­ising lobby is ex­pli­citly op­posed to set­ting browser de­faults for high­er levels of “Do Not Track” pri­vacy. There will be noth­ing to pro­tect them from un­wit­tingly giv­ing away vast amounts of data about who they are. 

On the oth­er hand, these are the tools that al­low web­sites to eke out a tiny bit more money than they oth­er­wise would. I am all too aware of how dif­fi­cult it is for me­dia busi­nesses to sur­vive in this new en­vir­on­ment. Sure, we could all throw up pay­walls and try to make a lot more money from a lot few­er read­ers. But that would des­troy what makes the web the unique re­source in hu­man his­tory that it is. I want to keep the In­ter­net healthy, which really does mean keep­ing money flow­ing from ad­vert­ising.

I wish there were more ob­vi­ous vil­lains in this story. The sav­ing grace may end up be­ing that as com­pan­ies go to more ob­trus­ive and high­er pro­duc­tion value ads, tar­get­ing may be­come in­ef­fect­ive. Avi Gold­farb of Rot­man School of Man­age­ment and Cath­er­ine Tuck­er of MIT’s Sloan School found last year that the big, ob­trus­ive ads that mar­keters love do not work bet­ter with tar­get­ing, but worse.

“Ads that match both web­site con­tent and are ob­trus­ive do worse at in­creas­ing pur­chase in­tent than ads that do only one or the oth­er,” they wrote in a 2011 Mar­ket­ing Sci­ence journ­al pa­per. “This fail­ure ap­pears to be re­lated to pri­vacy con­cerns: The neg­at­ive ef­fect of com­bin­ing tar­get­ing with ob­trus­ive­ness is strongest for people who re­fuse to give their in­come and for cat­egor­ies where pri­vacy mat­ters most.”

Per­haps there are nat­ur­al lim­its to what data tar­get­ing can do for ad­vert­isers and when we look back in 10 years at why data col­lec­tion prac­tices changed, it will not be be­cause of reg­u­la­tion or self-reg­u­la­tion or a user up­ris­ing. No, it will be be­cause the best ads could not be tar­geted. It will be be­cause the whole idea did not work and the best minds of the next gen­er­a­tion will turn their at­ten­tion to something else.

What We're Following See More »
Trump Ends Obama’s “Operation Choke Point”
36 minutes ago

"The Trump administration has ended Operation Choke Point, the anti-fraud initiative started under the Obama administration that many Republicans argued was used to target gun retailers and other businesses that Democrats found objectionable. Assistant Attorney General Stephen Boyd told GOP representatives in a Wednesday letter that the long-running program had ended, bringing a conclusion to a chapter in the Obama years that long provoked and angered conservatives who saw Choke Point as an extra-legal crackdown on politically disfavored groups."

Gorsuch to Deliver Speech at Trump Hotel
41 minutes ago

"Liberal groups are raising questions about a speaking appearance Supreme Court Justice Neil Gorsuch plans to make next month at the Trump International Hotel in Washington. Gorsuch is scheduled to headline a luncheon celebrating the 50th anniversary of conservative group The Fund for American Studies on September 28, days before the next SCOTUS term begins October 2. Steve Slattery, a spokesman for The Fund for American Studies, said Gorsuch had nothing to do with venue choice, which was made long before the group asked Gorsuch to speak."

Administration Faces Exodus of Top Cybersecurity Officials
1 hours ago

"The Trump administration has lost a handful of individuals serving in top cybersecurity roles across the federal government in recent weeks, even as it has struggled to fill high-ranking IT positions. The developments present hurdles for the new administration and speak to the longstanding challenge the federal government faces in competing with the private sector for top tech talent." Among those resigning is Richard Staropoli, "a former U.S. Secret Service agent who served as chief information officer (CIO) of the Department of Homeland Security for just three months," and Dave DeVries, the CIO at OPM. Separately, the White House announced today that President Trump has directed that United States Cyber Command be elevated to the status of a Unified Combatant Command focused on cyberspace operations.

Former Top Aide to McConnell Says GOPers Should Abandon Trump
22 hours ago
Trump Defends Confederate Statues in Tweetstorm
1 days ago

Welcome to National Journal!

You are currently accessing National Journal from IP access. Please login to access this feature. If you have any questions, please contact your Dedicated Advisor.