Google says once third-party cookies are toast, Chrome won't help ad networks track individuals around the web

Google says it will not come up with new ways to track individual netizens as they browse the web once Chrome phases out third-party cookies, commonly used for loosely observing people's online activities. In effect, the browser will not provide ad networks – and Google runs a very large one – alternative identifiers that can …

[Snake]

Targeting

Interesting that so many people want to discuss ad targeting but the elephant in the room, the very tech that is destroying our civilizations - SEARCH targeting - gets an open pass.

Thanks to Google Personalized Search, hundreds of millions of people are getting pulled into the rabbit hole of conspiracy theories, anti-vax, radicalized ideologies and violent behaviours. Just a few searches on any field brings you, and more importantly locks you, into an algorithmically-derived series of select answers...that the compose computer thinks you want to hear. Forget other views, other facts. Or any facts at all.

And it's destroying us. Reality is now what your search results can confirmation bias you to.

[Anonymous Coward]

Re: Targeting

@ Snake

How are civilisations being destroyed?

Your post stinks of "cancel culture"

Grow up

[sabroni]

Re: Your post stinks of "cancel culture", Grow up.

If you don't understand how this is destroying civilisations you could ask for an explanation.

Or you could shout your latest "get out of jail free" card ("brexit means brexit" is so 2010!) and pretend this is an attempt to silence something.

Is it Google's right to promote conspiracy theories you're worried about cancelling?

[Mast1]

Re: Targeting

Quite the opposite: "confirmation bias" means that expanding your horizons, eg participating in the debate, has been "cancelled".

[Dan 55]

Re: Targeting

The dangers were highlighted a decade ago in a 2011 Ted Talk by Eli Pariser (9 minutes), about 7 years after Google started personalising search and 5 years after Facebook personalised their news feed. The speaker argued that extreme personalisation based just on what the user likes seeing should stop or we risk dividing society and the audience, which included employees from Google and Facebook, gave a rapturous applause at the end.

And then nothing changed and here we are today.

[Mage]

Re: Targeting

And will Google stop tracking everyone via Chrome and all the Google services Web sites use?

Google doesn't need 3rd party cookies. Probably doesn't even need cookies, though logins use them for simplicity.

[Falmari]

Re: Targeting

@Dan 55 Thanks for that link

[tip pc]

Re: Targeting

I remember in the early days when a search for a specific term would bring back those results at the top of the first page, now I have to go digging down the middle or bottom of a page and wonder why my specific term was not at the top.

[Richard Jones 1]

Re: Targeting

At one time, if I wanted, a crosshead screwdriver, I would be offered, rubbish like 'The Hotel Cross Head Screwdriver'. At least that does not happen now.

The bane of my searches are those brain-dead 'comparison' sites that cannot offer you the damned item you want from a seller who could supply it in this lifetime. They only push the dross of their moment

[Christopher W.]

Re: Targeting

Snake's comment would be of legitimate concern, provided that Google's personalized search actually applied to search results on a broad scale. However, Google reports that personalization is only used for very small cases, such as current location for location-based results (i.e. weather and "near me" keywords), and bias for the embedded ad results. A quick search shows that other news groups have tested this and found it to be accurate. Further, a 2019 independent research paper has been done to conclusively prove exactly what Google claims - search personalization absolutely does not produce the bias that is being accused of in Snake's comment. https://link.springer.com/article/10.1140/epjds/s13688-019-0217-5

I hope others realize that conspiracy theories come in many variants, and we should always apply reasoning and research to our claims.

[Snake]

Re: filter bubble study

Thank you for the link :-) All that, and yet the study hasn't made any significant impact in the theory of filter bubbles in real-life application; indeed, it appears as if the "study" has been completely discounted.

As it should be.

The "study" was massively flawed. Major points:

1) No effort was made to study or discount the well known YouTube filter bubble, a bubble that has been well verified by users throughout the world;

2) the study used an opt-in plug-in for Chrome and Firefox. The MASSIVE errors of their implementation are:

a) all users were allowed to remain logged in to their online accounts, including Google;

b) no latent user browser data was demanded to be cleared, that is existing cookies were kept on all systems;

c) they used an installed plugin sample size of 4,384 - not exactly large as compared to worldwide internet usage;

d) all plugin installations occurred in one geopolitical region (!), and finally;

e) the plugin auto ran a background search of only 16 search terms.

So:

Any bias already set, to any user who volunteered for the study, was not reset due to that fact that clearing cookies was not demanded as a prerequisite to participate.

The users were not required to log out of, or prevented from using, any of their Google services. So the object in study - Google - was allowed to continue to openly interact with the users of the study without oversight of said study, including any possible butterfly effects said usage may incur.

The search terms were relegated to 16 political terms, "limited to the seven major parties and their respective party leaders", and *exclusively* inside Germany. But then they draw conclusions on Google search regionalization.

Flawed? I think you should have a good picture by now o.O All one has to *do* is look at your own Google search results...then clear your cookies, especially the persistent ones.

And let's not talk about your search return differences after clearing all your YouTube data O.O That's a completely different world you've just opened yourself back up to when you do that for yourself!

[Dan 55]

Re: Targeting

In addition to that mentioned above by Snake, this was just Google searches from 4000 people about the 2017 German election campagn, and even the paper admits as much "Thus the opportunity for the effect mentioned in Eli Pariser’s filter bubble theory to occur in this data is also very small".

So any extrapolation from that will probably come to the wrong conclusion.

[ZekeStone]

Once upon a time...

Once upon a time, you could set your browser to refuse all cookies or prompt you before accepting cookies.

I typically set it to prompt me to select between 'accept once', 'always accept', 'deny once' or 'always deny'. With that feature, it was interesting to see all the garbage some site would want to put on your system in the form of various cookies.

And that feature alone enabled me to avoid malware at least a few times

Then then browser makers made it harder and harder to do that before taking away the function completely... claiming that it 'didn't work anyway'... a claim I thought was bogus.

With some luck, the tide is turning and this functionality might come back eventually.

[tip pc]

Re: Once upon a time...

Firefox to the rescue.

[Dan 55]

Re: Once upon a time...

The amount of cookies being flung about on each individual page download made it impractical to continue having Yes/No pop-up for each individual cookie, but Firefox does offer cookie controls for many types of cookies based on what they are used for (tracking, social networks, etc...).

[iron]

Re: Once upon a time...

> Once upon a time, you could set your browser to refuse all cookies or prompt you before accepting cookies.

You still can. It never went away. It's called Firefox.

I'm amused by all this recent talk of banning third-party cookies and if it will break the web, I've had them banned since the 90s. Even IE4 would allow you to block all third-party cookies.

As for Google's new tracking method it's quite simple - Chrome will require a Google login before you surf and report all your browsing back to Papa Sundar at the end of your session. What could be easier?

[Snake]

Re: Once upon a time...

Same here. Firefox, cookies set to auto-delete upon exit plus blocking of all third-party cookies. NoScript. Privacy Badger. uBlock.

Firefox has always had, and kept, almost everything you need to control your own cookie privacy destiny, yet people choose Chrome instead.

[Someone Else]

Re: Once upon a time...

Firefox has always had, and kept, almost everything you need to control your own cookie privacy destiny, yet people choose Chrome instead.

Or "people" choose it for you; I speak from first-hand experience.

[Anonymous Coward]

Re: Once upon a time...

Or "people" choose it for you; I speak from first-hand experience.

Yep. In my case, "People" is what they officially call our HR department...

[Nate Amsden]

Re: Once upon a time...

firefox removed the ability to prompt to accept cookies a long time ago(I think it was just after firefox 33.something). I held onto it as long as I could then switched to Pale moon, who eventually I guess had to retire that feature probably a year or two ago(because of upstream changes). Tried waterfox back before I decided to use Palemoon but the feature did not work at all in that browser either at the time.

I still have 37k entries in my moz_perms sqlite database which I assume pale moon still uses(can right click on a page and see the permissions for that page and they seem to hold up), though don't have a way to add more entries(easily anyway).

$ sqlite3 permissions.sqlite

SQLite version 3.31.1 2020-01-27 19:55:54

Enter ".help" for usage hints.

sqlite> .schema

CREATE TABLE moz_hosts ( id INTEGER PRIMARY KEY,host TEXT,type TEXT,permission INTEGER,expireType INTEGER,expireTime INTEGER,modificationTime INTEGER,appId INTEGER,isInBrowserElement INTEGER);

CREATE TABLE IF NOT EXISTS "moz_perms" ( id INTEGER PRIMARY KEY,origin TEXT,type TEXT,permission INTEGER,expireType INTEGER,expireTime INTEGER,modificationTime INTEGER);

sqlite> select count(*) from moz_perms;

37009

sqlite> select * from moz_perms where origin like "%thereg%" limit 5;

35|http://www.theregister.co.uk|cookie|2|0|0|1512317577932

36|https://www.theregister.co.uk|cookie|2|0|0|1512317577932

37|http://nir.theregister.co.uk|cookie|2|0|0|1512317577932

38|https://nir.theregister.co.uk|cookie|2|0|0|1512317577932

130|http://forums.theregister.co.uk|cookie|8|0|0|1512317577932

but zero entries for theregister.com I guess they changed that after I lost access to that feature.

[heyrick]

Re: Once upon a time...

I use a combination of UBlock Origin to restrict what can run scripts and stuff, Cookie Autodelete to allow cookies and then delete them after 120 seconds (except for whitelisted sites), and a don't track me Google link unmunger to convert the ping-to-mothership search links into links that actually go where they say they go.

All runs happily on Firefox for Android (60.something, the slightly older not-broken one).

[Sparkus]

The unasked questions is

what has Google cooked up to replace cookie tracking to sustain their own ad business?

[sbt]

Re: what has Google cooked up to replace cookie tracking

Since it's only third-party cookies, not sure they need alternatives; the CNAME loophole to make third-party cookies first-party is going to be really hard to beat.

[A.P. Veening]

Re: what has Google cooked up to replace cookie tracking

PiHole can already blackhole CNAME DNS requests, making those cookies completely irrelevant.

[sbt]

PiHole

Two problems; it's not an industry wide solution; I'd wager less than 1% of internet connections have Pi-Hole protection. Secondly, blanket CNAME blackholing will break legitimate CNAME usages.

[A.P. Veening]

Re: PiHole

I am not advocating an industry wide solution, it is an easy individual, private solution. And PiHole doesn't blanket blackhole CNAMEs, only CNAMEs resolving to already blackholed sites/domains.

[sbt]

You're alright, Jack?

OK, well you put out 1% of the brush fire with your personal bucket of water. Problem solved!

[A.P. Veening]

Re: You're alright, Jack?

Every little bit helps and I put out 100% of that brush fire threatening me, so for me the problem is indeed solved and I am willing to share the solution with others. What have you done about it so far?

[Tony W]

Re: You're alright, Jack?

The solution is impractical for all but a completely insignificant number of people. Pi-hole is far from problem free, it might well require you to change the router which costs quite a lot of money for a good one and is by itself way beyond the ability of most users.

[A.P. Veening]

Re: You're alright, Jack?

It is one minor setting (which DNS server to use) in the router, which is usually provided by the ISP. And by now PiHole is problem free enough I recommend it to anybody asking about it. The only real obstacle is the investment in a bit of extra hardware if you aren't already running a server yourself, but an RPi should be cheap enough for just about anybody.

Having that PiHole use Unbound is only slightly more difficult and the major hurdle in that is knowing about Unbound.

[Christopher Reeve's Horse]

Re: You're alright, Jack?

Quite a lot of routers don't allow you to change the DNS address - all the Homehub devices provided by BT for example. Sure, there are workarounds such as setting up different DHCP servers, but it's less than ideal.

Either way, DNS content filtering is already on shaky ground with the advent of DoH, as the local client can specify whatever DNS it wants. At this point solving the filtering content becomes much more complicated than your average home user is going to be able to manage.

[Claptrap314]

Re: You're alright, Jack?

If a router does not allow you to change the DNS, then it is not fit for purpose. Return it, demand a full refund, and let everyone know about the situation.

Of course, if you missed the setting, the end result might be embarrassing/hilarious. I don't spend my time following the latest in routers, but I have a REALLY hard time seeing this actually happen.

[Getmo]

Re: You're alright, Jack?

The Pi hole software has a DHCP server built in. As long as whatever type of (crap) router you have that doesn't allow you to change DNS settings (I've never experienced this before) allows you to disable its DHCP server, it's almost trivial to set up.

Yeah it's true DoH is eroding some DNS filtering, but you'll notice that DoH and devices that ignore local network settings come from certain types of devices, usually of a particular brand. Like Chromecast dongles, Amazon Fire sticks, Samsung and Roku "smart" TVs, the usual suspects. You know, the behemoth data-collection players who have everything to lose if their slurp pipe gets pinched.

If you haven't installed any of those things in your home, it'll be much more effective. And if you have, many people don't connect them to the internet, or VLAN & firewall them into a corner. These companies' push for DoH is almost certainly due to the rise in popularity of pi-hole, and it's ease of use. Mine still regularly blocks between 35-60% of DNS requests, depending on browsing activity. DoH isn't as widespread as you claim, pi-hole still does good work, mine's blocked 51% today. And most importantly, it saves my smartphone battery life when browsing news sites, and removes most of the cancer they've developed.

[Anonymous Coward]

Re: You're alright, Jack?

I hope you know that more and more DNS will use DoH which as currently implemented uses application preferences and policies over anything the network 'suggests'. So unless you have configured all applications and somehow override all mechanisms that talk directly to DoH behind your back... then your DNS filter will become ever more ineffective.

[Getmo]

Re: You're alright, Jack?

DoH isn't that widespread yet, and so far it mostly only seems to be used by those devices & companies that have a vested interest in slurping your data: Google, Amazon, smart tvs, etc. I'd even say this push is due to the rising popularity of DNS filters like pi-hole and it's effectiveness.

Plus, I'm still not ruling out a potential backlash against these devices that don't honor local network settings. All it would take is for one of their products, like a streaming stick or smart TV, to be setup in the lobby of a business with a highly-secure network, like a bank, and someone to notice bank & customer data being illegally exfiltrated. Then it's billionaire companies vs billionaire companies, and massive potential lawsuits.

[IGotOut]

Re: PiHole

1%? I'd say it's a magnitude lower than that, more like. 0.0001%

[A.P. Veening]

Re: PiHole

I don't know the numbers, but by now setting up a PiHole is about as easy (and a lot quicker) than installing Windows, which has become easier over the years as well. And that RPi costs a lot less than a new Windows machine, leave alone a Mac, as well.

[Paul Crawford]

Re: PiHole

What proportion of the general public (not El Reg readers) install Windows? Being 10 or 100 times easier is still a very small number.

[David Woodhead]

Re: PiHole

I don't know the numbers, but by now setting up a PiHole is about as easy (and a lot quicker) than installing Windows, which has become easier over the years as well. And that RPi costs a lot less than a new Windows machine, leave alone a Mac, as well.

Virtually no-one installs Windows. It comes preinstalled on your PC or laptop. The concept of setting up PiHole, or a new Windows installation, or any reason for doing so, is so far beyond most people's comprehension as to be in the realm of magic. You may as well ask them to sacrifice a unicorn.

[vtcodger]

Re: The unasked questions is

what has Google cooked up to replace cookie tracking to sustain their own ad business?

Short term, they may not need to replace cookie tracking. They already know everything they care to know about you and me and most everyone else on the planet. Their competitors mostly know less. Bingo! Competetive advantage.

Of course long term their vast data store will go stale. But it's 2021 and who cares about the long term as long as the money flows copiously this quarter and next?

[sreynolds]

Re: The unasked questions is

I can see them starting to broker your information to third parties. Stop you from tracking people, but allow them to do it and sell a subset of that information back to you.

[Cuddles]

Re: The unasked questions is

Fingerprinting. Note the wording of the title - "Chrome won't help...". All they're saying is that they won't use the browser itself to do the job of tracking. Because they simply don't need to when there are plenty of perfectly good* methods to identify users without it.

* For certain values of "good".

[Kevin McMurtrie]

Re: The unasked questions is

It was probably completed years ago. They already have cloud backup of your Chrome history, subdomain web bug tricks, tracking JavaScript that every company embeds in their pages, and a Chrome update daemon that can help link together browsing history across multiple IP addresses just by having Chrome installed.

[Version 1.0]

Re: The unasked questions is

The answer is that Google now owns "advertising" and is making moves that will eliminate competition.

[Pseu Donyme]

Re: what has Google cooked up ...

One thing that could come handy is the data collected with Google Analytics. The key here are the 1st party id cookies that analytics.js generates and stores not only on the user's device (for two years) but on the Google Analytics server. While these apparently only identify a user in the context of each site, they still mean that when / if you revisit the same site (within two years) Google can be sure it is you even if your ip-address has changed. Otoh, if your ip-address has not changed since the last visit, Google can be sure* that ip-address == you between the first and second visit. This means that Google, in effect, has access to most everyone's** browsing history covering not only the sites using Google Analytics but also those using other Google services embedded in webpages (such as doubleclick.com, googletagmanager.com, fonts.googleapis.com/gstatic.com).

* most of the time, at least when the time between visits is shortish; otoh, given their huge data trove they can also tell with high confidence whether or not a given ip-address has been assigned to someone else in the meantime

** most everyone would be those who don't get rid of cookies on a regular basis i.e. most everyone

[Anonymous Coward]

Better than nothing? Maybe? Ideal? No.

If we take Google at its word (not always a great idea), this seems like a step in the right direction - towards less targeted advertising and less collection of PII.

Ideally, though, the only form of targeting that should be allowed is broad locations (like city or county) so that local businesses can avoid advertising to someone across the country from them. Unfortunately this will require government intervention through laws or regulations and I don't see that happening.

[Anonymous Coward]

Re: Better than nothing? Maybe? Ideal? No.

While laws and regulations might solve it, a bigger approach needs to be taken. Look at what some companies did to abide by GDPR or in some cases, circumvent GDPR. The laws and regulations need to be unified to a common set across the globe. Even just the major countries would suffice. If USA, Canada, Mexico, the EU, Britain, Australia and Japan all signed on, then chances are, the rest of the world would get the same treatment. Not much use having different systems and policies when most of the user-base/money all have common laws and regulations.

[Richocet]

Third party data brokers

The third party data market is a swamp. Data gets laundered so you don't know who collected it and anyone can buy personal data if they are prepared to pay. That means organised crime, the Kremlin, stalkers, fraudsters, anyone can access it without identifying themselves. So I will be happy to see that go away. Strengthening Google and Facebook's position for advertising is a bad side-effect though.

[sabroni]

re: it's not clear exactly how this will impact Google

It won't. Chrome will keep on sending each users unique product id to Google for them to monetise as they see fit.

Good job they're not M$, eh? We dodged a bullet with that IE crap!!

[Anonymous Coward]

Re: re: it's not clear exactly how this will impact Google

Well, at least Chrome is less malware-prone than IE. That IS a slight benefit.

(Written in Firefox)

[big_D]

CNAME

Doesn't CNAME collusion automatically open up the host site to a GDPR violation, because all cookies are then 1st party, which means the marketing company can look up all that juicy personal information in the host site's cookie, heck, they could misuse the session ID to browse the site in the user's name, if they really wanted to.

[Mast1]

".........product management in Google's Ads Privacy and Trust group..."

Am I just too old & cynical in thinking that this had an echo of "Newspeak" ?

[Anonymous Coward]

sounds like the guy had a touch of 'Chris Grayling' and got promoted sideways into a job at the end of a long corridor in the basement of an annex at the back of Big G's campus... possibly with 'Beware of the Leopard' on the door

'"Privacy & Trust Group", you say... nope, no idea... have you tried asking the guy over there with the mop and bucket if he knows where it could be?'

[Graham 32]

They work just fine. I've had 3rd party cookies disabled for a few years.

[iron]

No they don't. I've blocked all 3rd party cookies since the 90s, most sites work fine and I temporarily enable them for those that don't maybe twice a month. Less than once a week anyway.

Of course I don't use FB, Twitter, etc so YMMV.

[tiggity]

Non targeted ads are better

IMO

If I'm on a retail website looking at loudspeakers, then targeted ads, showing me loudspeakers on offer at some other site might be of interest.as it relates to what I am looking at

A "targeted" ad, showing me washing machines (because 2 weeks ago my ancient washing machine died so needed replacing) is utterly pointless because I got a new washing machine ASAP as its not a device I want to be without for long. A washing machine ad on day I was browsing for them would have been fine, 2 weeks later a total waste of time. Targeted ads just do not have the level of nuance required... whereas ads related to current content have a better hope of relevance.

[iron]

Re: Non targeted ads are better

Showing you ads for speakers when on a speaker website IS targetted ads. Just using a different metric to target you by.

[Anonymous Coward]

Re: Non targeted ads are better

I like targeted advertising. It allows me to decorate all websites I visit with images pleasing to my eyes, rather than random crap.

[Anonymous Coward]

Re: Non targeted ads are better

How do you get all of your ads to be rewritten to load random images from your personal library? That would be an extension I'd pay for.

[Old Tom]

René Magritte

Recently I couldn't remember René Magritte's name; a google search or two later and my memory was jogged, case over. Except that I then see adverts featuring some of his work for weeks afterwards.

'Targetted Advertising' is complete rubbish.

(And we've all had those repeated adverts for something we've already bought or that we never wanted to buy but recently researched.)

[Anonymous Coward]

Re: René Magritte

just a few days ago, browsing on my phone, I got hit with a spate of ads for flats in a monstrosity being built next to where I work!

I don't have GPS switched on unless I really need it ('Google Maps' opens at the last place I searched and history is off, Chrome/Google search lists me random places in SE England), I don't live locally (25 mile commute), I don't browse/search for local places or businesses (not on the phone anyway) and I'm certainly not looking for a place to live, so I guess the only way they could pinpoint me would be from sniffing wifi... either that or it was a very remarkable coincidence

(that was with Chrome, I've since installed Firefox)

[Anonymous Coward]

@AC - Re: René Magritte

On top of the protection measures you mention, I never browse the Internet on my phone so Google can safely insert the vast amount of data collected somewhere into their Southern end of their back.

All they have is my desktop browser on my PC (not Chrome). As for Gmail, I open it exclusively into a portable version of a different browser, again not Chrome. So they probably have disparate pieces of data and a limited way to reach me to monetize it.

However, I am aware that Google is still making money selling those pieces of data.

[Alumoi]

Re: René Magritte

But you have bluetooth enabled, right?

[Unicornpiss]

If someone really wants to improve the world

..Then making telemarketing of all kinds (including spam texts) illegal would go a long way towards that, IMO. Or make it too expensive for the fly-by-night scammers to afford doing it, such as a mandatory charge of say, 5 cents a call and extreme penalties for anyone trying to avoid this or violating do not call lists. Making spoofing CLID (caller ID info) illegal would be great too. I know this would be very nearly unenforceable, but a man can dream..

There are few things that irk me more than walking through a cloud of gnats in the summer or getting some idiotic "extended warranty" call with the ID spoofed to look like a local number. And it always happens when you're incredibly busy and/or waiting for an important call. No one ever answers their phones any more and this is why.

[Anonymous Coward]

Re: If someone really wants to improve the world

In the US, caller ID spoofing *IS* illegal. Doesn't slow down these scammers, though. I wish my phone company would provide the ANI number instead, as that would be much harder to fake.

[stiine]

Re: If someone really wants to improve the world

Its only illegal if its done with intent to defraud. Companies can still rewrite CLID so that all outbound calls come from their toll-free number, or any other number they control.

[Anonymous Coward]

Re: If someone really wants to improve the world

True - which clearly makes the scam calls' use of CLID spoofing illegal.

[Anonymous Coward]

What bugs me

Is not so much the targeted ads, since apparently my browser is already blocking around 2,500 trackers

It's ad networks, deliberately used by websites, that fill the page with "Houseowners in [location] do this" or "How to cure [thing] with [thing]"

[Anonymous Coward]

Modern advertising doesn't work anyway

Yes, I'm well aware that there are "studies" showing it works - but I seriously don't believe it. The vast majority of ads I see are for products/services I wouldn't buy, and no ad can change that. For those I would consider buying, the ad only serves to let me know of a company that sells it.

Seriously - when was the last time you clicked an ad? On purpose, I mean.

[Anonymous Coward]

@AC - Re: Modern advertising doesn't work anyway

Don't let them fool you! They are making insane amounts of money out of it, so for them it is working perfectly, irrespective of you clicking or not on the ads they shove down your throat.

Remember, advertising industry makes money on the promise that those ads will be thrown at you. It doesn't matter at all if you watch them or not because they've already got paid.

[low_resolution_foxxes]

Ad tech companies are one virus in this nasty swamp.

Idiot marketing execs are worse. I occasionally speak to our marketing team monitoring our website and it is frightening how much granularity is available on who is viewing our site. I'm not sure precisely how it does it (no login functionality) but the marketing droids think it is perfectly normal to see user names, emails, company, location, company details for a moderate cross section of our users.

Granted this could just be those that filled in the "Contact Us / newsletter" details and based on ip address maybe?

Without this data the marketing droids will have less info and will panic (a very good thing IMO). They will likely try to migrate to whatever will make them look good in front of management ("marketing insights!" as they call it). A website marketer without any info is useless

[Anonymous Coward]

'erosion of trust' – gee, wonder who could be responsible for that...

... 99.99% of people, who shrug and click, regardless.

[DS999]

Google doesn't need cookies

They control the browser the majority of PC and nearly 100% of Android users are running, which can track people FAR more effectively than cookies.