> Only about 9% of Safari users on an iPhone allow outside companies to track where they go on the web
> In comparison, 79% of people who use Google’s Chrome browser allow advertisers to track their browsing habits on mobile devices through cookies.
Something to keep in mind the next time you see someone from Google complaining about how Apple won’t allow Google Chrome to run natively on the iPhone.
Apple will let whatever browsers want to run on iOS. Chrome is, in fact, on there.
They just won't let them using any rendering engine but WebKit—and this restriction long predates Google's current level of draconian dominance in that sphere.
I don't believe the tracking settings are tied to the rendering engine, so I would be surprised if Chrome on iOS has settings that different to Chrome on macOS in that regard.
The rendering engine also includes the entire network infrastructure. I can't think of how I would plug my own network infrastructure into WKWebView on iOS, which means I can't block any built-in tracking prevention.
I think you’re confusing WKWebView with SFSafariViewController. Content blockers are not used for WKWebView. You can use your own content blocking logic via the appropriate delegate calls, but the “system-wide” content blockers are not inherited from Safari.
I hate tracking like everyone else but what you wrote is just pure FUD.
Apple forces Firefox/chrome to use their crippled engine but that has nothing to do with the parts that do the tracking.
For the record, Firefox has built-in anti tracker which is enabled by default and they do that on all platforms. And don't forgot the privacy enhancements Samsung added to their browser years ago.
Using “FUD” in one sentence and calling WebKit “crippled” in the next is not a good way to establish yourself as a rational analyst. Firefox is my primary browser but I don’t think advocacy like that is helpful.
This hasn’t been true for years. Yes in the past, the embedded safari engine (used by apps) didn’t have JIT capabilities. But since webkit2’s multi process model, and WKwebView in iOS which is now the only way to embed WebKit, it’s the exact same as native.
That doesn't feel like the right word: it's mandatory but you're still talking about a very competitive browser engine and the security rationale for blocking third-party JIT is not without merit even if I wish they'd relax it[1]. This feels similar to the IE 5-6 era: it got really bad by the end after Microsoft removed most of the team but there was an earlier period where they were forcing it on users but most developers would have acknowledged it having good performance and features relative to the competition.
In this case, it's especially complicated by the fact that iOS is the primary check on web standards meaning whatever Google chooses to implement. Mozilla isn't without influence but there's no incremental way to chip away at Apple's market-share the way they have with Google/Youtube.com pushing Chrome at Firefox users short of making Android phones far more competitive.
1. Say by requiring strong commitments to code quality and response time, which would allow e.g. Firefox and Chrome but not Adobe-level security commitments.
A forced choice is only a severe limitation if it’s significantly worse. That’s currently a guess as there are no iOS comparable options to consider at this point in time.
Further, Google has a competitive interest in eventually crippling their iOS rendering engine.Then again I don’t use desktop Crome so that could be my bias talking.
No, they wouldn't ever do that. That would be like making an anti-demo. Google has an interest in getting as big a market share for their browser as they can. In as much as disabling their web application functionality of you do not use chrome.
Android is much less relevant than the browser.
And this does not even begin to touch on Firefox.
The problem is a policy one, with Apple really liking their walled garden and veto rights to any design.
For sake of argument lets’s suppose Android Crome was a 90 and iOS Crome is currently a 70 due to the rendering engine. In theory they could replace the engine and hit say 85 given enough resources. But, they could also replace the engine hit 75 and blame iOS. Their still demonstrating an improvement and intentionally crippling it at the same time.
That said any such measurement is arbitrary. I personally like having a unified rendering engine because it’s less testing when developing mobile websites. There is no way to improve that situation by introducing a new engine.
Technically they are not introducing a new engine. Chrome is well established with a 60-70% market share. Not saying thats a good thing either...
But you have a very pro-apple and anti-google way of thinking here which I think misguided and does not actually benefit you. Even with those made up numbers chrome would improve from 70 to 75 which is by all means a win for the consumer (you!) but you are ready to dismiss it as it might be bad for Apple.
I edited the post to make it clear I don’t think Crome has a better engine having seen many issues on android phones. I was simply saying if they had advantages on other platforms you can’t assume they will carry over.
Anyway, I still test and find different issues with Crome on Android tablets, phones, and desktop due to significant market share and I would do that for iOS Crome. Sure, Chromes developer tools helps fixing some of these but I have had various CSS issues not replicate.
However, most companies don’t do this so adding more rendering engines with slight bugs seems to make things worse for consumers. There may be short term benefits, but long term it’s a net loss as many developers would only support a subset of iOS users.
> 3rd party browsers are forced to use a webkit engine instead of their own. By any definition, this adds a severe limitation => crippled.
Not only that but if WebKit has an exploit, all browsers have an exploit.
> I also think (I could be wrong) safari has access to a slightly different engine than third party browsers.
Correct, Chrome used to be based off WebKit's HTML rendering engine, which Google forked and created Blink instead. Firefox has their own engine (Gecko?) which is all being redone in Rust.
> For the record, Firefox has built-in anti tracker which is enabled by default and they do that on all platforms
Firefox's anti-tracking protection doesn't block google analytics by default. They only block "third party" cookies - which in practice translates as: Google's competition.
The fox knows better than to bite the hand that feeds it.
The above specifically shows you to make sure your site continues to function when Google Analytics is blocked by Firefox. I can't find any evidence that Firefox doesn't block Google Analytics.
Mainly 1st party but some features might use 3rd party.
--
For customers that are using Google Analytics' Display Advertiser features, such as remarketing, a third-party DoubleClick cookie is used in addition to the other cookies described in this document for just these features. For more information about this cookie, visit the Google Advertising Privacy FAQ.
And when there is a security vulnerability which exploits webkit by just visiting a website, one has to wait till Apple releases an OS update(if it decides to acknowledge and fix it[1]), where as one on Firefox 'if Apple had allowed its own engine' could get updated via the AppStore.
What Apple has done with the trackers is commendable, but the security argument when it comes to browsers is always in favour of allowing browsers with their own engines; which apple doesn't.
I may be misremembering but my understanding was that Apple disallowed other rendering engines not primarily for security reasons (although I'm sure they're a factor), but for performance and, by proxy, battery life.
It's not dissimilar to the argument against Flash: yes it was insecure and unstable, but mostly it was a performance hog and a battery drain, which isn't great on mobile devices with limited battery life.
What’s the practical difference between an OS update that can be one tiny patch and an app update? Not only does iOS 13 support all phones back to 2015. They released a patch for older devices going back to 2011 less than 6 months ago.
The reference does not back up your claim - Apple never disputed that there was an issue, and they had already fixed it when Project Zero published their report.
iOS Firefox uses Webkit, not Gecko as intended. It does link to your Firefox accounts and so on but otherwise it is much harder to distinguish against other iOS browsers.
The browser rendering engine used has nothing to do with tracking permissions, which is set by the browser itself. Chrome is on iOS for example, as is Firefox.
From Modern Day's Startup CEO, CTO or SVP that are under 30 or barely 30+ getting their 5 minutes of frame, while never actually programmed in the IE era and wrongly assumed IE era as IE 7 or IE 8.
As someone who worked in ad tech for half a decade, I fail to understand the fuss about cookie based targeting. We're speaking of databases mapping data to random numbers stored in a file that can easily be deleted. The real issue is ad targeting by a handful of large companies that have personally identifiable information.
It's not like cookie based targeting is very effective either. It's very difficult to get cookie based ad targeting right – you have to make significant investments into product, technical integrations and data buys. Given the fact that the cookies to which you are tying these investments can vanish at any time (and a large % do so regularly), it rarely makes business sense to make the right sort of investments. Instead most companies do the bare minimum and try to make a guess that is slightly better than a coin flip. I actually remember seeing a deck from a data company boasting how their gender data was ~ 60% accurate.
Why are we worried about this? Contextual targeting is far more effective. As an example, targeting an ad on Elle.com is more than 60% likely to reach women than an ad targeted based on cookie data.
Genuinely curious so would appreciate calm, non-aggressive responses. I'm wondering whether this is due to a generally poor understanding of cookie targeting effectiveness or if I'm missing something.
You probably would be well advised to read up on the economics of consumer differentiation and price discrimination.
All of today's ad-tech is set up and developed for that purpose. Whether overtly or covertly, it is the singular purpose, because the case is so clear cut: We are distributing the share of the pie between companies and consumers.
It is therefore clear why people aware of basic economics consider agressive tracking and privacy violations hostile, because they ultimately are or will be.
I am not worried about actually showing ads that are personalized in some way. But as others have pointed out, cookies are not for that. You can target based on the site's content and you are probably showing me a much more relevant ad anyway.
I am worried that every offsite ressource and every ad is used to track me across hundreds or thousands of shady services that operate under zero regulation or control, the combination of which allows market actors to determine how much of my consumer surplus they can extract in every sense, and thus actively devaluing the internet for me, in excess of what happens in the real world!
And the responses of the ad industry lobby groups pretty much prove to me what they think about choice and what they expect to get from consumers.
At this stage, it is entirely rational and ethical to consider the entire ad industry as hostile.
As a group, with all side-effects, the notion of just "showing an interesting ad" simply does not exist anymore. At least for the collective industry. Any effort you took to track users, no matter how noble, will ultimately be hostile to me in the future.
>All of today's ad-tech is set up and developed for that purpose. Whether overtly or covertly, it is the singular purpose, because the case is so clear cut: We are distributing the share of the pie between companies and consumers.
That's not entirely true though. Better ad targeting increases the "size of the pie" by increasing the chances that a consumer will find something that's useful to them.
>At this stage, it is entirely rational and ethical to consider the entire ad industry as hostile.
At that point, why not consider the entire software industry as hostile? They're the ones that enable this, so clearly they are also the devil. Life is not this black and white. Ads do serve a useful purpose, to which degree cookie based ads do that, I'm unsure, but it would be very difficult to start a business without any kind of advertisement.
>That's not entirely true though. Better ad targeting increases the "size of the pie" by increasing the chances that a consumer will find something that's useful to them.
Most ads people see are not for novel products solving problems. The ads are just fighting for brand recognition and space in peoples brains. This parasitic and wasteful industry is squandering some of the brightest minds of several generations to dare. Ad cos are also wasting untold billions of dollars suffocating development of infrastructure and industries that increase quality of life.
> [...] the combination of which allows market actors to determine how much of my consumer surplus they can extract in every sense
That would only be possible if the online stores could perfectly discriminate users based on price. Which happens to some extent (booking flights is a popular example) but is difficult to pull off and also illegal in most countries. You are not allowed to make personal discounts for each customer for example. Better and stricter legislation to prevent these kinds of personalized discounts would be pretty straightforward to implement if it were a major concern.
Here's a recent example from the EU: https://eur-lex.europa.eu/eli/reg/2018/302/oj It addresses geo-blocking customers of online shops (and redirecting them to national versions of the shop with different prices).
Price discrimination itself is not illegal (for example in the EU), it is only so if it is based on certain parameters. And this you'd need to prove in court, which makes it unlikely that regulation will solve this issue any time soon.
Beyond the legality of the issue, the point remains: Here, companies and consumers are strategically at odds.
Why is price discrimination bad? It is about charging the highest price the customer is still ok with. Sure, if a merchant knows how badly you want his product he can increase the price. But, on the other hand, he will decrease the price for somebody who wouldn't pay the high price.
It's like bargaining on a fair. A method to determine a price suitable for both sides - instead of charging every customer the same price.
And it's not (only) about how much money the customer has. A rich customer can still be price sensitive and would get a low price based on price discrimination.
Also, in my opinion, ads and tracking are (mostly) orthogonal concepts. A website can track or show ads or both or do nothing. Tracking is often used for ads but not exclusively so.
Sometimes price discrimination may be overall good, but it easily leads to price gouging and I believe there a laws against it, for a good reason.
Should I pay 1.2x, 3x or 100x as much today, just because you know I desperately need the product/service today vs. yesterday?
What happens when I get (wrongly) classified as rich and now can't afford a normal life?
Should medication cost <som of your savings> if it is for life threatening issues? Should basic necessities cost more because my house burned down/got robbed/lost my luggage/whatever and you know I have money?
The argument against that is that if sellers cannot increase prices in times of crisis they have no incentive to stockpile for such events. This leads to shortages and even higher prices on an unregulated black-market.
I'm not saying that I buy this line of reasoning, but there's the counter-argument.
It's not really an applicable argument here. That's a general time/availability/word situation discrimination, which I think I would be ok with, to a limited degree.
I was arguing against people based discrimination, without any meaningful relevance on availability. Again, I think reasonable and reasonably qualified discounts are ok, but not increasing prices for the same product and service.
At best, price discrimination is a waste of time and resources. The buyer loses as much as the seller gains, making it a zero-sum game. But since price discrimination requires more effort than setting a single fixed price, the sum is actually negative.
Aside from that, "charging the highest price the customer is still ok with" means to eliminate the consumer surplus. Which is bad for consumers, i.e. everyone.
And aren't prices supposed to be signals? How are customers supposed to compare the costs of each good if you don't tell them? All you tell them is how much they're willing to spend.
They already know that. But you (probably) don't. If you did, congratulations: you've solved the economic calculation problem. This should raise red flags.
> But, on the other hand, he will decrease the price for somebody who wouldn't pay the high price.
I highly doubt that they would actually decrease the price. They maybe wouldnt increase it, but unlike the bargaining on a fair, the seller can not see that the prospective buyer is really struggling to make ends meet, and sell it at a loss to help them out.
They don’t normally decrease the price below what the original baseline was. Would be nice if they did, but that’s not how it appears to be used by today’s companies
>My browser history is none of your business. That’s it.
Your browser was made to handle cookie data. It is publicly known. If it bothers you, you can turn it off. I do not see what all this sort of huffing is about.
If you walk in to Walmart you don't automatically think you're sharing your shopping habits with thousands of other companies who pay for that data[1]. People believe the web works the same way - when they visit someDomain.com they, entirely reasonably, accept they're sharing their data with someDomain but don't realise they're also sharing their data with whichever tracking/surveillance services someDomain has installed.
You're right that this is just how browsers and websites and the internet as a whole works, but what tech workers and tech companies are just recently discovering is that people don't actually like or trust how it works and aren't always willing to accept it. In Apple's case they can leverage that distrust to market their products.
[1] Today facial recognition and phone tracking means that you are 'sharing' that data with tracking companies, obviously.
> If you walk in to Walmart you don't automatically think you're sharing your shopping habits with thousands of other companies who pay for that data
You can literally buy that data. Either by targeting credit card transactions of people who shopped at a walmart via a broker or you can get more direct data (albeit probably more limited) from them directly here: https://www.walmartmedia.com.
That's the entire problem! Every store doing this should have a mandatory seminar before every shopping visit that outlines every single data point they collect, names every company, intermediary, and consumer of the data they collect directly and clearly, one. by one. These parasitic and disturbing methods only thrive because they have not been brought into the light of day so every consumer can consciously choose what data they want to share at every point. Any other methods except complete informed consent are fraudulent, invasive, and should be illegal. The sheer greed and lack of any morals in ad tech is sickening.
>If you walk in to Walmart you don't automatically think you're sharing your shopping habits with thousands of other companies who pay for that data.[1]
Isn't that exactly what membership cards for stores are for? Isn't credit card data also sold in a similar way? I find the credit card one to be a much bigger problem.
Consider this: the vast majority of people you targeted more than likely didn't understand the terms of the deal (buried in ToS, hidden behind flowery language like "personalization", didn't understand the scope of such cookies, etc).
That is not informed consent.
What you were doing was exploiting people by going beyond the scope of their knowledge of the world and rationalizing it by working backwards from "well, they must understand what's going on" to "they must like it or get something out of it."
Problem is that's not true.
The reality is, even today barely any of them likely have the faintest clue what you did to them.
If they knew they would likely be far less than thrilled.
> It's not like cookie based targeting is very effective either. It's very difficult to get cookie based ad targeting right
Tell that to the likes of Amazon and Ebay. Their cookie-based advertising is extremely effective. They not only get to advertise things at you that they already know you like, they also learn about other things you like by knowing what websites and content you're viewing.
Why are we worried about this? Because megacorps learning everything about us isn't desirable. It only serves to intensify the advantage they have over individual sellers, and in doing so, lessening competition over many layers of business (all the way from manufacture to delivery).
I'll agree these things apply much less at smaller scale, but that's also why this is so important. Handing the biggest marketing advantage to the biggest players isn't good for anybody but them.
Amazon, Ebay, Walmart, etc. have persistent cookies. They are mapped to your profile in their database. If you delete the cookie, it reappears the next time you sign in. These companies can collect vast amounts of data on you, which allows their targeting to become highly effective over time.
But these companies are few and far between. Most ad tech companies have nowhere near this quality of data or access to users.
if they do that it's very, very subtle. I get mostly product that I bought, with a handful of product that I saw recently, and zero original suggestions, even if I'm open to impulse buy, I am driven by new impulses, showing me something I haven't bought once is not going to tip me over as that decision was already rationalized.
In the adverts they show you what is most likely to convert today. Stuff you've been looking at recently. But every adview is another blob of data to profile you with for future recommendations.
We talk about Facebook knowing too much about us but they really only have a social handle on us. Amazon knows where we live, what we buy, how much money we have, when we're going on holiday, when we're pregnant, how many children we have, what pets we have, what cars we drive, what we like to watch, what we like to listen to, when we're home.
Every new B2C service they launch is another intrusion. Advertising is just one ring in a long chain. It gives them reach over everything you view online.
Good point, for me it’s linking my cyberspace identity to my physical identity and having that sold on, stolen aggregated. Increasing my security vulnerability to phishing and identity theft
So you would agree if I store a cookie with a (randomly generated) ID in your browser to remember what you liked, what you bought, how you interacted with my website etc.? Also, I can sell this data to third parties and show personalized ads - all without knowing who you are outside of the web..
Probably this. Adtech (ironically) do not advertise what they do. Most people read and think that e.g. google literally sells personal files on themselves.
Another 5 year adtech worker here (and I'm still in it): it doesn't matter that it's a handful of large companies that have PII, it's that those companies keep getting more and more of adbudgets, and continue to build more and more products towards continuing to build real profiles.
"In 2016, Internet advertising revenues in the United States surpassed those of cable television and broadcast television. In 2017, Internet advertising revenues in the United States totaled $83.0 billion, a 14% increase over the $72.50 billion in revenues in 2016"
The fact that I worked in ad tech doesn't automatically make an advocate for ad targeting. I don't like being tracked either.
As to your question on why it's being done if it isn't effective, I believe there are two reasons for it:
1. A vast % of digital ad spend goes to a few companies that have personally identifiable information and very effective tracking. Google, FB. I did point of this exception in my comment. Frankly, these companies don't need to target you to be effective. They gather their richest data from your usage on their service.
2. As for the rest of ad tech: Media buyers at agencies tend to be fresh out of college grads. They don't fully grasp the capabilities/limitations of ad tech products, targeting effectiveness, etc. Agencies want to spend their digital budgets and not exclusively on FB, Google. They want to show that they're diversifying, innovating, etc. This IMO is how the rest of ad tech stays afloat.
That's not addressing the argument. Advertising follows eyeballs, so naturally it went online. There is no evidence that its effectiveness has increased however, despite all the claims about high ad tech.
> There is no evidence that its effectiveness has increased however
If you're right, those who commission ads are idiots because they aren't doing the most basic cost/benefit analysis.
And this in an area where doing so is very easy (don't run the ad then do run the ad, then see the immediate and longer-term difference).
Also, from the same wiki link I gave
"The EU limitations restrict targeting by online advertisers; researchers have estimated online advertising effectiveness decreases on average by around 65% in Europe relative to the rest of the world."
Further, by "There is no evidence that its effectiveness has increased" are you saying no studies have been done, or studies have been done but show no such evidence? In the latter case please post a few links, thanks.
> those who commission ads are idiots because they aren't doing the most basic cost/benefit analysis
i should clarify i meant the effectiveness of ads overall, not just targetting. there is no doubt that online ads work, and that tracking ads perform better than display ads. However despite the shift to online ads, advertising spending overall is similar and it doesnt seem to have increased consumption. So i dont understand if the rise in online ad spending means anything (imho it's purely because online audience is bigger now).
It used to be that Flash-based cookies were pretty solid, but now there’s no more flash...
What matters to ad campaigns is lift and not accuracy. What about conversion attribution? Don’t you need cookies for Amazon/etc to figure out what website led you to a purchase?
What the article lacks is a breakdown of who the ad buyers are. In particular, there’s probably a big difference between today’s Facebook vs display ad buyers (especially vs 5 years ago).
Cookie based targeting is relative easy to remove. It is an early and easy vulnerability that can be quickly addressed before the big ones that takes more engineering and fundamental changes.
Contextual targeting is a bigg problem, and the fix will require a fundamental change. It is almost impossible to use technology to prevent contextual targeting in areas like email service and social media as long as the data is in clear text and available to the server. The current best bet is to use laws like GDPR but it is slow and not something a browser can dictate with a software update.
Is this a serious question? Just because you want to target something doesn't mean it is ok to do it. Geez, the arrogance. You want to do X, so the rest of us have to give up our privacy? Technical feasibility is not the same thing as ethical, courteous, or just plain being respectful of other people. The whole industry seems to have done some sort of massive mob rationalization of inappropriate behavior. It's not ok.
If I am on a tech site reading about disk drive failure rates, feed me an ad about an SSD sale. If I am reading about mountain climbing, show me ads about climbing equipment. Fine. But don't start following me around the web with adds about ascenders for the rest of the afternoon. Or travel insurance. Or ... It is none of your business what I browsed half an hour earlier.
Thank goodness at least one business (Apple) has enough customer focus left to try to address some of this nonsense. All the better that it helps them against some of their competitors. That just makes them more motivated. Bully for that. At least we are still the customers rather than the products with Apple.
Couldn't agree more. I hate it when people respond to concerns with "I make money that way, how could I possibly go on otherwise" as if it somehow legitimitized their unscrupulous activity. Their profits are nobody's problem but theirs and there is absolutely nothing wrong with putting them out of business if that's what it takes to stop their abusive surveillance capitalism. If people's use of privacy-focused software and products reduces their revenue, so be it.
I was just asking the question because I was genuinely curious whether there was technically a way to do so w/o cookie targeting. I have no interest in doing such targeting. I was interested in whether a HNer had a creative workaround.
The way you asked the question suggested you were trying to find alternative methods to track people. That implies the method of tracking (cookies) was the problem rather than the tracking itself. My response to that is tracking people isn't something that should be done to begin with. People should not want to track others. It should not be possible to track people in any way, let alone make any money from the practice.
I'm not sure there is any stronger interpretation available here. You're baselessly speculating on gp's emotional state. That is the opposite of "assume good faith."
I actually agree with your point. I was just asking the question because I was genuinely curious whether there was technically a way to do so w/o cookie targeting. I have no interest in doing such targeting. I was interested in whether a HNer had a creative workaround.
> I was interested in whether a HNer had a creative workaround
There is still fingerprinting. If you take that away also then it gets trickier and fuzzier. But hey, if I can recognize you by the way how you walk then it should be even easier to do so by the way how you move the cursor or how you scroll. Or the time you usually visit my website. IP-address? Ping-times, animation delays.. And if somebody pays me good money I will find even more opportunities :-P
> What if I want to target women who recently bought Chanel perfume? How would I go about doing that unless I do video surveillance and face-recognition targeting?
Suppose a woman walked into a dress shop, and the clerk said, "We find that our $LINE line of dresses over there works really well with the Chanel perfume you just bought."
Do I have to explain why 1) people think it's creepy, and 2) why people think you don't have a right to do that?
If those women gave you permission to contact them, that’s how. If someone bought Chanel, you don’t have a right to that information unless they explicitly gave you that right.
What about the retailer that sold the perfume? They'll usually auto sign up the customer to their product emails and/or adapt their landing page to customer's interests (like Amazon).
Under GDPR this kind of self advertising falls under legitimate use of an email address, so if you inform the user you'll send them the emails, it's legal to do so without explicit opt-in. (AFAIK IANAL)
> The cost of reaching Safari users has fallen over 60% in the past two years, according to data from ad tech firm Rubicon Project. Meanwhile ad prices on Google’s Chrome browser have risen slightly.
The demand for highly-targeted users is what drives up prices. Since Safari users all look the same now and can't be highly targeted, they have lost that high-end demand and prices have fallen to generic display rates.
I would disagree. "People who can afford Apple gadgets" is a desirable high-income audience. I wouldn't be surprised if it was broken out as a special category for purchase by FB and GOOG soon.
The point is they are less valuable because you can't track them, so you can't take advantage of their potential value.
Lets say an average Android user is worth 100 value units, and you can get lots of uniquely identifying information about them, say 90% of the ideal amount of info, so their effective value to you is 90 value units each (90% of 100).
Uniquely identifiable Apple users info is potentially much more valuable, say 200 value units. However you can only get a very small amount of actionable information about them because they are so hard to track, due to Apple's security and privacy systems. So maybe you can only get 10% of the information you would ideally like to have about them. That makes their actionable value to you only 20 value units each (10% of 200).
> The point is they are less valuable because you can't track them, so you can't take advantage of their potential value.
This is simply bullshit. So-called “targeted” advertising is a giant scam pulled on non-technical marketers who don’t understand it at all and use “measurements” provided by the ad networks themselves.
Supply and demand have stayed the same in quantity, but the quality of both has dropped severely. It's the same equilibrium but at a much lower price point.
Let's say someone is trying to sell snowboards using ads. Previously they could target say "young male users who have previously browsed snow sports." Let's say the ad had a 5% conversion rate and they were willing to spend $100 per conversion (sale). This means they are willing to pay $5 per ad displayed. Now they can only target "all safari users" and their conversion rate is now 0.5% since not everyone likes snowboarding. They are still only willing to pay $100 per conversion, but now they are only willing to pay $0.50 per ad displayed since they need to display 10x as many ads per conversion. If the auction rate is >$0.50, they stop buying ads completely which decreases aggregate demand for ad slots.
In theory a "user" is more valuable the more you know about them because the extra info lets you better target them for higher response rates. Better privacy means you know less about the user, meaning you are less likely to bid high for that user's eyeballs, meaning the user is worth less ad spend.
The actual story is a little more nuanced: when you target an ad, you are usually bidding for users in an auction. You're generally willing to pay more for more specific filters because you're more confident of a higher response rate.
In ecosystems with a lot of data on every user, specific filters actually work, so you end up with higher bids on fewer users, which should drive the average spend per user up without necessarily driving total spend up. With less data, you have to use less specific filters, so you end up with lower bids on more users, which drives average spend per user down without necessarily driving total spend down.
There's also a countervailing force, depending on the market. First an example that supports what everyone is talking about:
* User X loves cola
Pepsi and Coke both bid more to serve their cola subscription services or whatever. Ad price goes up. Now add more data:
* User is a Coca Cola fanatic.
Even though this is the same user, the ad price goes down, because Pepsi doesn't want to bid.
Now add another Pepsi fanatic for symmetry, and you can see how "more data about users for better targetting" can in some cases reduce auction prices (while of course still increasing advertiser value.)
But if users were trending towards either side then the number of users in the middle would presumably be diminishing, driving up prices to target them even more.
Your scenario just seems to single out one particular example where more information can reduce spending on one group, while increasing it on another.
I think this is only on fake-scenario. In a real DMP, the user would be "loves cola","loves coca cola","loves fizzy drinks","drinks sugared drinks", and about 30 other related segments, and one of those would definitely sneak by.
Humans create audience segments, and it's in the interest of dataset producers to have each user be a part of the maximum number of members they can be to maximize monetisation.
Eventually there will be machine learning models involved (maybe there are already? I don't know ads.) Those models might move closer to "perfect targeting" where advertisers could compete less over certain users.
And no matter the incentives of the data providers, "increase customer value" is probably always a Nash equilibrium. If better targeting leads to less revenue, eventually you'll probably do it anyway.
Imagine you have two users. One who likes apples and one who likes bananas. Both like oranges. I sell apples and am willing to pay $1 per conversion. You sell bananas and are willing to spend the same. OP sells oranges and is willing to spend $0.75 per conversion.
If we can't tell which user is which, we know 50% of our views will lead to a conversion, so you and I will both spend $0.50 per view. The orange seller beats us and the price is $0.75 per user (assuming an old fashioned auction).
If we can tell which user's which, I can spend my $1 on the apple eater, and you can spend yours on the banana eater. Both of us will spend more than the orange seller, who is still only willing to spend $0.75 per view. The price rises to $1 per user.
Dollar per conversion is the constant, and the pile of converting users doesn't change. With less targeting, there's just more hay in the haystack. Dollars per viewer goes up as the number of viewers goes down.
This makes it clear why it's in the individual advertiser's interest _and_ advertising network/website's interest to get more data on the users.
The cost to the advertiser has stayed the same (or in reality, has probably gone down slightly, as it's likely fewer people are bidding on the impression) while the revenue per impression has overall gone up.
Pricing on targeted ads is linked to confidence that a person will perform an action (buy, sign-up, etc). This is a function of the "richness" of data collected -- does the data about you contain demographics, engagement, behavior, so forth.
So a shoe company may pay a lot for a "male/18-35/basketball-fan/shoe-blog-reader". But Safari does not easily allow for that richness of data to be collected, so the price is low.
It's still trivial to target users where browser="safari".
But with Safari's recent changes you can no longer target users on Safari where "non-empty cart"=true, because that relied on cookies and tracking which are no longer available.
On a per-impression basis advertisers are willing to pay (way, way) more if they know they are reaching someone who has already visited their site and been engaged enough to add-to-cart. There are plenty of other examples but abandoned-cart-retargeting is one of the most obvious ways advertisers can see good ROI on expensive ad impressions.
You have to look at it as the companies buying ads pay more for users that are more likely to result in sales.
Exposure increases this value just by playing a numbers game; therefore, the larger the ad network, the more they can charge.
Targeting increases it more, by allowing companies to choose demographics more likely to be interested in their product, so again, they can charge more.
An interesting aside: While working as a mobile developer, I learned that many apps do not have an in-app-purchase to eliminate ads because it tanks the value of the in-app advertising. Why? Because the people most likely to drop $5 or $10 to not look at ads? They're the ones with the most expendable income, and therefore, users more likely to impulse purchase products advertised.
In other words: By allowing people willing to spend money to support the app, even if there is NO OTHER EVIDENCE they'd be interested in a product, decreases the value of your advertising.
I imagine it’s not that it’s harder to target whether someone is using Safari, but that it’s harder/less effective to try targeting demographics within the group of Safari users.
Advertisers look for a certain ROI on their ads. If Safari users are harder to target, then the absolute return on ad spend will be lower (apparently enough to offset the affluence of Safari users). The way to bring the ROI of Safari ads in line with the ROI of ads on other browsers, then, is to pay less for Safari ads.
Are ads a lot worse on Safari as a result of this price drop? I’ve tried going tracker-less before (e.g. on YouTube) and the ads get hyper spammy. There was a large shift, too, in the spamminess of display ads a couple years ago as Facebook took all the high-quality ad dollars. There are several news sites that run such spammy ads they usually crash the browser.
Sometimes I wonder if the entire ads racket isn't a case of the emperor with no clothes.
Has anyone here actually had a meaningful increase in users and revenue by using targeted-ad services, that they couldn't achieve by more organic methods (e.g. posting on places like Reddit, word of mouth)?
The real goal for the big players using these services isn't actually increasing their own users and revenue. They have better ways to do that.
Rather, the goal is to lock everyone else out from even possibly doing so.
For example: AdWords isn't about reaching new audiences, it's about staking a claim on an audience in perpetuity (as long as your budget allows). This is completely apparent from the pricing model.
Google's brilliance is that via the ubiquity of their search platform, they have effectively managed to leverage a property tax on attentional real estate. If you don't pay, someone else will move in and hopefully develop the attentional property further, raising the value Google can charge you for it until you're priced out and someone else comes along to hold onto it.
That's like the 1000th ad-targeting article I've read, but the first time to hear this perspective. Absolutely fascinating, thank you!
Here's one question though: If you're the established player in one field and don't really need the online ads other than for staking a claim, would it really influence your sales that much if a competitor takes over that space? Probably just a downturn in number of sales to new customers? (Who presumably haven't heard about the other players in the field and just go with the one that's shown to them in online search results?)
Okay, that's more than one question, but maybe you'd like to elaborate a bit.
I very much doubt the small bump in sales of Coke after a major ad campaign actually pays for the campaign. That's not the point.
Lets suppose CocaCola stopped advertising completely. No TV adds, no product placement, no billboards, nothing, but Pepsi and the other soft drinks companies kept up their rate of advertising or even increased it to fill the gap. The only time anyone ever saw the name Coke would be when they saw it on a shelf. It might take a while, but the outcome would be inevitable.
It's known for quite a while that only a minuscule fraction[1] of the digital advertising results in actual conversion. But, marketers have come into terms that even it is worth spending large money; it's just that only those without a large budget and expecting miracles out of digital advertising end up as total losers.
Then again, with the influencers/celebrity culture, the landscape of digital advertising has changed. Hence, Instagram supposedly leads when it comes to organic conversions. Good news here is, those who don't want such ads can stay away from the platform (or) that influencer. Bad news is, it is resulting in creating hoards of gullible, programmable masses with little effort, who can be exploited for various nefarious activities which can have an effect even outside of those platforms.
Retargeting can be quite effective, along with other forms of first-party data used for targeting- such as CRM data to inform which ads are relevant.
Behavioral data using 3rd parties like Neustar / Lotame have more dubious results, but still rake in billions each year from ad buyers that fastidiously track their ROI.
Most online display works to reinforce messaging in other forms of media, or win over new audiences with sheer volume.
Good question, this [] article proposes that this is indeed a bubble. I tend to agree and generally avoid ad-tech related work, even though it is closely tied to the engineering work I do.
Personal experience shows that poor results from targeted ads is usually because of a) poor product, b) poor audience targeting, c) poor creative, or d) poor testing and tracking.
There's an entire cottage industry of marketers (especially affiliate marketers) who make a lot of money through targeted ads. Unlike big brands, they can't afford to throw money at branded campaigns; they have to see results. If they continue to patronize ads, it means that they are getting conversions.
Advertisers don't care and have never worried about this. It's the adtech supply chain, and the publishers who get paid that worry about the revenue and targeting.
Depends on the advertiser! Big brand who mostly wants to reach everyone in a geographic area? Targeting doesn't matter. Retailer that wants to bring someone back in to complete a purchase? Without targeting they won't buy ads at all.
You suggest this as if the majority of ad spending comes from small businesses trying to retarget their customers; when the vast, vast majority of ad spending comes from big brands; and your mom and pop business isn't running a sophisticated ad targeting campaign.
How has this been so effective, when Panopticlick still says my Safari browser has "a nearly-unique fingerprint"? (And based on what I've read from other users's experience, this is normal.)
When you load Panopticlick, you'll notice it takes a minute or so to fingerprint you. That's a pretty long time and a lot of CPU.
Re/Targeted ads require the decision to be made in milliseconds, including network transit time. They just don't have the time to fingerprint you on every request. They could do it once, but then they are back to the problem of having to set a cookie to keep track of you after that.
>When you load Panopticlick, you'll notice it takes a minute or so to fingerprint you.
No, it takes a minute because it’s trying to load trackers in order to test your adblocker. Fingerprinting itself takes only a moment. Try TorZillaPrint: the results are instantaneous.
> Re/Targeted ads require the decision to be made in milliseconds, including network transit time
If they're served on page-load when a user first visits a site, but I guess a significant minority are also requested+served in the middle of your newspaper article two minutes after page load, or in the middle of your Youtube video, or a refreshing sticky-footer ad, or just the next page loaded on a site.
It's mobile Safari that is hard to fingerprint and getting harder with each version. Most adtech still runs off cookies to share data and store identifiers and those have been basically eradicated.
Having to fingerprint every pageview also creates latency and still isn't that deterministic so everything will need to shift to probabilistic and contextual targeting. That's a hard change to make for most legacy networks.
Telecoms like Verizon and ATT have their own adtech companies and were doing this, but regulations and encryption technologies have disabled most of it.
IPv6 with probabilistic modeling comes pretty close though.
Fingerprinting is much harder in reality than Panopticlick makes you believe.
For example, if you update your browser or add a new plugin, you can look like a new person so it makes it harder to find a match. A cookie would not change through that event.
Enough people have the same fingerprint that it limits its effectiveness. Maybe you can only reliably identify 40% of your user base that way.
It is like using your IP address as an identifier. For some people it can be 100% accurate, but others switch networks or share an IP with multiple other people which limits its usefulness.
What ends up happening is that companies will correlate IP address with other fingerprint information to get a probabilistic match but that is complex and still has flaws.
They say "nearly-unique" - I get an odds of '1 in 50,000'. Granted, that may be enough combined with IP or some other factor as a transient/session cookie. However the hash might change as soon as a plugin or some dependency is updated.
Combine my Canvas fingerprint (which a plugin update won't change) with my rough location (based on IP address), and I have to imagine you'd get very, very close to identifying me.
Canvas fingerprints are probably the same between other iphones of the same model and ios version. So unless you use a rare model and/or rare ios version, there's probably at least a few dozen people with the same fingerprint/ip range combination as you.
My understanding—only because another HN commenter told me as much, many months back—was that even the exact same model hardware would return different canvas fingerprints due to silicon differences. Not true?
That seems implausible unless there's analog logic involved. With digital logic a bit is either a 1 or a 0, and any "variations" (eg. a 0 where it was supposed to be a 1) would be considered flaws as they can dramatically skew the end result (eg. the MSB of an integer being flipped, or one of the bits in the exponent of a float). As far as I know, there aren't any analog logic involved in a rasterization pipeline.
That sounds unlikely. I'm sure there is some variation due to differences in OSes and GPUs, but there are surely enough iPhones with each to make it nothing more than supplementary entropy.
you can be in California and using say an iPhoneX and it will tell you you're one in 500k. If you calculate the number of people with their iPhone set to English in the PST timezone (the only things there is to distinguish one iPhoneX from another) and then you divide that by the pantoptclick number you get it's claiming there are only 10-100 iPhoneXs in all of Seattle Metro Area, San Francisco Bay Area, Los Angeles Metro Area, and San Diego Metro Area combined. Clearly wrong
> Your browser fingerprint appears to be unique among the 251,543 tested in the past 45 days.
It only counts against those other people who visited panopticlick, and that in the last 45 days. That's explicitly stated, but hidden in the click-through for details section.
There are also things it can't detect, such as the existence of the Cookie Auto-Delete extension. So it thinks cookies are enabled and can be used for tracking, but in reality they get deleted whenever I close a tab (and links to new domains get opened in new tabs). Or user-agent randomization, which will make your fingerprint more unique but less useful for tracking.
I'd love to know this too--maybe fingerprinting isn't yet being used most places? Or maybe fingerprinting isn't that useful without additional cookie tracking?
From what I understand, a lot of the ad networks still work off cookies, they haven't moved to using fingerprinting yet. I think as browsers implement more anti-tracking features they may start to though.
Not sure if this is the case but it would be possible to fool fingerprinting by randomizing parts of the browser so each request has a unique fingerprint which isn't useful for tracking.
Related. The next big change for privacy and ad targeting is the removal of the IDFA by Apple and the GAID by Google. Removal or aggressive randomization, but there is no way those IDs remain as persistent as currently.
It's a matter of months, maybe a couple of years.
When this happens, what is happening with Safari now will look like a minor issue.
Yes indeed. And reset your GAID when you want and know how to do it. But my point is that such a behavior will be forced by the OS rather than offered as a choice.
It's what is happening with Safari (and the other browsers). First, knowledgeable users are empowered to make a decision, then, that decision is rolled out automatically to the masses. E.g., https.
Let us not forget that Apple isn't doing this out of altruism, but 50% because it hurts their competitors and 50% because it makes their users happy. That said, being 50% motivated by making their users happy is more than can be said about Google or Facebook.
Edit: To be clear, I am an Apple customer and it's mainly on privacy grounds. I just like to keep realistic expectations about their present and future behavior, instead of buying into Tim Cook's rhetoric of idealism.
It doesn't matter that a business isn't making a business decision based on altruism. What matters is that the incentives of the users are aligned with those of the business
I think it's considerably better when it's not based on altruism, because it's easier to predict how stable or sustainable an alignment of interests might be.
At an Apple all-hands around 2005, an employee asked Steve why they made the iPod, and he said "because we love music." And the answer was so simple and so earnest, that you just knew it was true.
Apple, for better or for worse, makes products based on what its executives personally want to use. That's not the only consideration, but it's unusually salient compared to other companies.
My guess is that Apple works to limit ad tracking because its executives personally do not want to be tracked.
Apple under Steve Jobs 2005 and Apple under Tim Cook now are different beasts. Steve would go crazy at the executives releasing all the buggy/unpolished products Apple released recently. Imagine the Macbook keyboard or the Touchbar being released under Steve.
iPod was a great product. But time has changed, and I would be cautious in extrapolating that experience to Apple today. I can't imagine Apple executives saying "We change the keyboard because we love typing".
Apple had plenty of misfires under Steve as well. iTools, Apple Hi-Fi, etc. People forget.
The crappy MBP keyboard illustrates my point. I suspect that Apple neglected the Mac for a long time (in many ways) because its executives switched to using primarily iPads [1].
I feel the same way. Imagine how Steve Jobs would feel if he found out you could go to an Apple store, buy a Macbook and an iPhone, and still have to buy an additional cable to connect them together!
It's be less ridiculous if the iphone didn't come with a USB cable. It did, though, just the "wrong" one (I understand from elsewhere in the thread they've finally fixed that, however).
"It is not from the benevolence of the butcher, the brewer, or the baker that we expect our dinner, but from their regard to their own interest." - Adam Smith
I like the term that Ben Thompson (who writes Stratechery) uses for this kind of thing: "strategy credit":
> Strategy Credit: An uncomplicated decision that makes a company look good relative to other companies who face much more significant trade-offs. For example, Android being open source
Agree -- total red herring. And I'm much happier that Apple has a financial incentive to keep this up. "Corporate altruism" has a tendency to go poof! the instant it conflicts with the bottom line.
One can make the argument that Google's "do no evil" was pure altruism. And indeed, it did go poof! the instant it conflicted with their bottom line. Astute comment, parent.
These days they try hard to sell themselves as "the good guys of tech". Tim Cook grand-stands about privacy rights, etc. It's easy to forget that it's all just business, so it's good to be reminded.
>That said, being 50% motivated by making their users happy is more than can be said about Google or Facebook.
They all have products to sell to users so in that sense they all have to strive to make those users happy. Some business strategies use lock in to keep consumers but unless you want to wage that accusation against one of these three companies it seems like they should all be equally motivated by consumer happiness.
The closed/open/free/freedom paradox of Apple is something people should be thinking more about:
They are in many ways the most closed of all the big tech companies. Their flagship platform (in revenue terms), iOS, is the most closed general purpose computing platform in history. Overall it's second only to things like game consoles for being locked down to everyone including the owner.
They are also the platform and OS that offers the best privacy and security to the average user, and the only mainstream platform/OS actively pushing back in any way against surveillance capitalism.
Why is this? Three reasons I think, listed in order of importance:
(1) Economics: Apple's closed premium brand gives them pricing power, causing the majority of their revenue to come directly from their users. This makes it at least economically possible for Apple to prioritize their users' privacy and security at the expense of other potential secondhand sources of revenue. If you're playing in the bargain bin or worse are "free" then you must seek revenue elsewhere, and monetizing users via surveillance and other scummy methods is the easiest and most lucrative way to do that.
(2) Security: being unbelievably locked down is a double edged sword: it robs users of choice, but also robs malware and adware of attack vectors. Android is more open and as a result is a spyware disaster and seems to be getting worse. There's so much un-policed attack surface that Android's permissions don't really mean very much.
(3) Corporate DNA: Apple still has some DNA left over from the personal computing era back when PCs were about empowering users rather than monetizing them and the Internet was supposed to be about communication rather than manipulation. Apple still thinks it's making products for people, not productizing people.
> iOS, is the most closed general purpose computing platform
> in history. Overall it's second only to things like game
> consoles for being locked down to everyone including the
> owner.
It's fair game, that is why competition is healthy. If other companies don't try other formulas you end up with having all tech companies be ran by ad revenue.
people are too prone to anthropomorphizing companies. capitalism in two words: incentive alignment. people should really never be expecting business decisions to be motivated by morals.
People are going to unpleasantly surprised by the way the internet changes after tracking for personalized advertising stops working. Already, I see tons of complains about paywalled news sites, micropayments everywhere, and the death of the "old internet". Just imagine how much worse it'll be when only two or three giant social media companies can make money with advertising (since they can rely on first-party data) and everyone else is relegated to a high-friction ghetto. If you're sad about the death of the "old internet", look out, because you haven't seen anything yet.
As for tracking: the current "privacy" push looks like a moral panic to me. I've yet to see any actual personal harm come to anyone as a result of cookie-based tracking. I have, on the other hand, seen billions of people benefit from the monetarily-free services that personalized ads enable. Whipping up an anti-cookie hysteria is making the internet worse for everyone but the big established giant companies.
It will likely make fake news less profitable, since I would bet that a lot of value in advertising on those sites is driven by the audience, not the type of content being posted.
I think that would be a nice return to the "old internet".
Also, it isn't clear to me that the larger sites you talk about would not recoup more value from their audiences and thus need to paywall less.
I've yet to see any harm from that lion hiding in tall grass over there, stop panicking and get back to grazing.
I have yet to see any harm from eating badly and not excercising for this young person, stop panicking and let them enjoy their double-upsized extra-bacon Monsterbuger.
All of that cookie blocking makes for smaller, and sometimes fewer requests. In total that represents a non-zero effect on battery life (and bandwidth) that Apple gets to use to bolster their battery benchmarks.
> For instance, a publisher can track what people do on their own site—just not on other sites. So they can sell targeted ads based on the information they have.
This is what I believe is fair. Unfortunately we are seeing the difference between first-party cookies and third-party cookies gradually vanish. Publishers are already encouraged by ad tech companies to set up CNAME records so that a subdomain on a publisher site points to an ad-tech company domain.
That may help the one adtech company but they still wouldn't be able to "sync" (https://www.adpushup.com/blog/cookie-syncing/) their cookies with other companies in the supply chain.
If so, that could mean fewer iOS apps will get written.
If not, is that only on technical grounds (it may be easier to hide what data apps send out) or (also) because it is good for Apple if one has to write an iOS apps to better track users?
I don't think it does and, as an advertiser, people mostly exclude apps from a lot of display/video campaigns at the moment but it's being touted as the next big opportunity because of moves like this.
iOS apps have an advertising identifier that can be reset by the user.
There are of course other ways to fingerprint the devices, but they will get your app removed from the App Store if you're using them for advertising. (If you're caught of course.)
The difference is that users can go to privacy, and switch a single option and they are opted out.
In Android, and with Google in general, you have to go through many different menus and hidden option in different places. And then, you can't really stop tracking either, because you'd need to also go to several of these "ad alliance" scam websites and try to disable tracking for hundreds of services. And of course, these websites never actually work for more than a couple of those.....
From my perspective, coming from Android, iOS is really good.
I am soo jaded by Google and Android that I just expect a company to try to obfusicate options and generally be hostile to my privacy.
I regret using Android for many years, especially since Google, while I wasn't lookin, quietly moved from innovative and user-friendly to literally user hostile.
> A Criteo spokeswoman said that by making the ad-blocking feature automatic in Safari, Apple “does not truly promote choice for the users of its browser.”
People don't go to the grocery store to give their money to the cashiers. So from a user standpoint the cash registers shouldn't be there in the first place.
The ads typically are paying for what the users are going to the website for. I don't like it, but most users seem to prefer visiting sites with ads over sites they have to pay directly for. I hate how intrusive and abusive many ads can be, block ads on many sites, pay to subscribe to some sites, and hope other models succeed someday, but have to acknowledge that ads do serve a purpose.
> Their proceeds allow websites to exist and motive entrepreneurs to create them in the first place.
The desperation about monetising ads usually makes them crappy website creators. They would just fill pages with low quality articles propped up by SEO. I prefer the sources that don't earn money on advertising for quality of content.
Everything Criteo says is just terrible. They're about as borderline legal as it gets. I'm 100% expecting them to get massive GDPR related fines in 2020.
It's a paywalled site. They're not tracking you other than to ask you to pay for the content. This will only become more common as sites shift from ad revenue to subscriptions.
I’m not too sure about that. There’s no guarantee they won’t both get your payment and then track you (with confirmed personal details you provided during payment).
I was talking to my friend about this the other day. He's an executive at an ad company (not Google). He had an interesting take on this:
Yes, it is in fact working and breaks targeting and retargeting. But what is the outcome of that?
The CPMs (how much the website makes) are 10x higher for re/targeted ads. So as they stop working, do you think the website will cut their revenue? No, they'll either put up a paywall, or put 10x more ads on their page. And those ads will frustrate you even more, because they aren't targeted to you.
And in the meantime, Google and Facebook will still know everything about you. It's just the smaller websites who suffer by not getting ad revenue. And they weren't even getting your personal data in the first place.
I still side with Apple on this one, but he has a good point. Overall what they are doing will probably just make things worse for everyone that isn't Apple, Google, and Facebook.
I'm not nearly as anti-ad as many people I've interacted with on HN and Twitter. I like the idea of getting free-to-me content supported by ads. I resisted running ad blockers for a very, very long time.
But.
The ads just got bigger. And bigger. And more intrusive. They approached 50% of the visible content (and that's on the desktop, not mobile). They played more video. They played more music. They started blocking the content I was trying to read, all in an effort to get me to click.
So I gave in. I run an ad-blocker, too.
Your friend inadvertently pinned what I personally think is the original sin of online advertising: measuring an ad's success through the number of clicks it gets. Ads in other mediums aren't designed with the idea that the only successful ad is one that makes you go AAAAH MUST GO BUY RIGHT NOW RIGHT NOW DROP EVERYTHING AAAAAAAA... but 20 years ago, some dingdong decided that because it was possible to measure "direct engagement" with ads on the web in a way you couldn't with TV or print, it was appropriate to measure ads that way.
Think of how different the web might be if, instead, we'd just sold ads by the number of visits to an article the ad is on... or better yet, by the number of visits to the web site, period. "But that's not an exact measure! We can't tell if it's really worth what we're being charged!" You also have no way of knowing whether or not that full-page color ad in the New Yorker is "really worth" what you're being charged, but you know what? You got along fine that way for decades. You did. Your business was fine.
You omitted the most important reason why blocking is necessary: malicious and deceptive ads.
I want to check a story by my local newspaper. I don't have my phone handy so I borrow the nearest work computer that isn't typically used for web browsing. I open an incognito window and go to the website, click on the article, and read through it. I leave the page open while I go back to work. Sometime while I'm not paying attention one of the many ads on the page refreshed because when I turned back instead of the newspaper I see a full-screen message that my computer has been "hacked". There's an pseudo-unclosable alert box warning me not to turn off the computer and to call "Microsoft" at some phone number. This computer didn't have speakers attached, but if it did I suspect they would have been making noise or using text-to-speech to read the fake warning.
This is a daily occurrence for thousands of unwitting computer users. You can't blame the victim by saying they were doing anything to make themselves vulnerable. I saw this happen on a continuously patched and vulnerability-scanned Windows 10, running the most up-to-date chrome. While visiting the website of a mainstream web site operated by a major newspaper syndicate.
I put the responsibility to stop this on the ad networks, and the site operators who partner with them. Until you can guarantee that none of the ads you serve are malicious or deceptive, I'm blocking everything and denying your revenue stream.
Me too. I worked for a company that made almost all of it's revenue from advertising. It felt hypocritical to run an ad-blocker. But it just got so bad I had to do it.
I'm just worried about what will happen to the web for all the people who don't run ad-blockers. I care about them too, especially since they are the customers in a lot of cases, and if they go away from the web completely because it's so bad, what then?
> Think of how different the web might be if, instead, we'd just sold ads by the number of visits to an article the ad is on... or better yet, by the number of visits to the web site, period.
That's actually how many ads are sold. By CPM (cost per thousand impressions).
I've heard the ad industry has been moving toward that model rather than clicks, which seems good -- but it doesn't seem to be appreciably helping make the ads suck less. For years Daring Fireball paid its bills and then some with a simple, non-moving, square sidebar ad; obviously that wouldn't keep the lights on at Vox, let alone the NYT, but I can't help but think ads could have been... better.
There are places that at least strike a balance that doesn't make me want to poke my eyes out -- The Outline, for instance, at least now that they've toned down the over-engineered JavaScript woo they launched with -- but it's depressing how uncommon that is.
> the original sin of online advertising: measuring an ad's success through the number of clicks it gets
The measure is often the number (or even the value) of conversions an ad gets. I think it is a good measure for everybody. I wouldn't block ads if they were garanteed to have conversion-rates > 90%. This would mean that I only see ads which are highly relevant. What we have instead are conversion rates < 5% (or even < 1%) - which is basically noise or spam.
>only successful ad is one that makes you go AAAAH MUST GO BUY RIGHT NOW RIGHT NOW DROP EVERYTHING AAAAAAAA...
Exactly. That would be a good ad. Instead we see useless distracting spam ads which bring no value to users nor businesses.
I don’t get targeted ads. They’re quite dumb. If I look at product X, ads for product X follow me around the web, which isn’t that useful because I’ve already done my research and am not quite ready to hit “buy” no matter how many ads you show.
Ads for similar products I might be interested in, but I’ve not seen too many of those served up by these ad networks.
Untargeted ads at least might show me something new and interesting.
That's not how they work. What you're seeing is the issue with not having data shared between networks and the resulting poor targeting.
Retargeting has better results than any other type of targeting but unfortunately there's a gap between the ecommerce systems knowing you bought something and the ad networks used to market those ads. It's easy to drop a pixel and say you looked at a product, but it's much harder to let that same network know that you're no longer interested for whatever reason.
Add in complexities like multiple devices, browsers, ecommerce stores, payment methods, and it only gets harder. This is usually why retargeting is much better with smaller direct-to-consumer companies which tend to have a single customer database and payment method to sync from.
Amazon has been repeatedly showing me office chair casters as a suggestion after I bought a set on Amazon.
Despite having much better information on things I actually spend money on than the average ad network (all my devices have adblockers), Amazon isn't clever enough to figure out that I'm probably done with that product category for years.
To explain how this happens will take several pages of notes on the complexities of ad serving that are the exact opposite of "not clever enough".
A summary is that your adblocking may have an impact on syncing stats, your purchases are not unique enough to matter for that category, serving some ad is better than serving no ad, and it maybe better to show you that impression however worthless than to bother with something else considering the auction dynamics at that point in time.
Do you think that's true when it comes to Amazon's recommendations, which I'm assuming are largely based on activity on Amazon itself? My adblockers don't keep Amazon from tracking things I do on their site while logged in, nor from showing me product recommendations.
Actually before I was ad blocking, two large local electronics stores did the same stupidity. I bought some expensive electronic toy, then was followed for a week by google ads for the exact same electronic toy. Which i had just bought. And was expensive enough to not buy another. And the ads were for the EXACT SAME STORE i had just bought it from.
I repeat, that was before I began blocking ads. For a while I was on the 'let the web sites make some money' side. Then the web sites became unreadable with ads on so it's uBlock Origin now.
That's what I described about the ecommerce system being different than the ad network. Products usually have their own pages where you can easily be added to a 'interested' segment.
It's very hard to remove you from that segment on a consolidated check out page that would need some backend mapping with the products you bought and the removal pixels for each product and ad network, along with lots of other issues. It's just not that simple as you may think to keep things in sync.
So they're okay with me considering them stupid? :)
As the customer, I don't care how hard or easy it is. I can imagine how it would be implemented, but 99% of the world population can't.
Note that they never gave me ads for stuff i only looked at, only for stuff that went through my shopping cart. If it's already there, I guess it makes figuring out if i bought it or not easier?
The explanation I've seen is that yes, most people who buy certain products (a microwave oven, say) are probably done for a long time and less likely than the average person to buy one, but overall the group of people who just bought a microwave oven are still much more likely than the average person to buy another one. So yeah there are a bunch of people for whom the ads seem extremely stupid because of course you don't need another one, but enough who think "you know, I could use one in (the camper/the basement/the rental/the garage-office), actually, but it'll need to be a smaller one" that it's worth advertising other options to the whole group. (for the record, I hate all ads and wish they'd go away)
I've read loads of suggested explanations for why this is a thing over the years, and none of them have ever felt satisfactory to me.
However, none of the people doing this really care why this is a thing, they care that, for some reason, this is a thing. It's vastly more profitable than nearly every other type of advertising, and since nobody _really_ knows how that works either, they just follow their tried and tested patterns that work.
Obvious _you_ wouldn't be this dumb. But it turns out that, on average, you are exactly this dumb ...
I suppose it might be true that the sort of person who buys 50 sets of office chair casters first buys one set to try out, and if they're happy with that, comes back for more in a week. The chance that I might be that person is relatively high reward (a ~$1000 sale), and Amazon doesn't have a better idea of what I'm going to buy.
The simple answer is that there's a far greater number of people who saw the product but didn't buy then those who did. Conversions are a tiny percentage that scales with price, so showing ads to everyone who saw it raises that conversion rate.
It doesn't really affect the people who already bought it since the conversion is complete, and the wasted impressions are trivial compared to the cost of implementing more complex tech to keep things in sync.
Perhaps there's a secondary effect of negative reputation from seeing ads for stuff you bought but the data doesn't bear it out.
>"but it's much harder to let that same network know that you're no longer interested for whatever reason."
Yes and no. As a best practice, for a SaaS product I marketed I would always build exclusion lists (ideally ones that auto-updated based on syncing with our funnel data) to stop showing ads trying to get people to buy who were already paid subscribers.
eCommerce obviously has different challenges, and the effort to create those synced exclusion lists may not actually be worth the performance improvement at scale. But it is certainly possible when an advertiser has both sales data and ownership of the retargeting lists.
I actually find myself clicking on targeted ads occasionally because they are so relevant. In that sense, I'd say it's the content that makes it less frustrating.
Your proposed tradeoff space are not the only two options.
As a ‘third way’, you can match ads to real content, and have audience interests and advertiser interests aligned.
Folks who try this, find it works, in some cases only a percentage point or two worse than high end retargeting, and in most cases, better than retargeting when sold by third party or retargeting across run of site.
EDIT: CPC (cost-per-click) and CPM (cost-per-thousand impressions) are both fine when matching ads and content. For instance, the ‘affiliate link’ business is predicated on the idea that inline clicks on products mentioned in content are so valuable as to be individually worth a piece of the actual sale.
> As a ‘third way’, you can match ads to real content, and have audience interests and advertiser interests aligned.
Doesn't work because the advertisers usually want to sell something different from what the majority of users are interested in. If they strictly respected the interests of the users most of the ads would disappear.
affiliate sites are created purposefully with persuasive text specifically to rank high on google and cause sales. in the generic case, a website would have to specify its own advertising-keywords, because contextual mismatching the audience.
Regardless, i think an ad service with only contextual ads will either fail, or end up implementing the wheel (targeted ads) to survive. tracking adtech does lead to more effective, less intrusive ads in the average case. The tech that will dethrone google will be a service that matches buyers and sellers better than google.
Apple is not willing to understand the whole dynamic in the ads market, including publishers trying to game with the system all time. I've observed the worst on what those publishers can do and am pretty sure about that they will find out a "creative" way to make up for the revenue loss. Of course, I'm not saying this will be user-friendly; publishers tend to be more "creative" when it tries to rather exploit users.
Do you work for/with Apple regard to this issue? I followed discussions with WebKit devs and had an impression that they don't really care about the ecosystem but only Apple users. It seems they had never reached out to publishers/advertisers in the design phase of ITP. Later, they prepared a crappy conversion tracking mechanism to compensate the loss, but this design also had never involved the ecosystem.
I do not currently work for Apple on this, but you can read about the ways Intelligent Tracking Prevention has changed to adapt to the ecosystem on the WebKit blog: https://webkit.org/blog/category/privacy/. I was under the impression that the ad click attribution work (which I think is what you’re talking about) is being proposed as a standard and is open for comments right now.
I understand how ITP works, its intention and goals since this had a significant impact on my work. And assuming a good intention on Apple's side, this makes me pretty confident that Apple doesn't want to put significant investment of its resource on studying how the ads ecosystem works (and of course they don't have a strong motivation to do so).
On browser-side conversion tracking, I can say that this will unlikely be adopted by advertisers/publishers in the current form. This is not my own opinion, but more about general feeling across advertisers/publishers that I've met.
apple has millions of devs of apps who amount essentially to an army of gig workers who expect to be paid via ads or in-app-payments. if apple is not making sure the devs have a fair revenue stream , it's equivalent to abuse of the same kinds Uber is accused of. sadly, the terms of all similar platforms favor exclusively the companies and not the devs. Still, labor laws exist.
Personally, I think I'm fine with that, provided the paywalls get cheaper—right now, they're all very clearly more expensive than what you'd make in ad revenue. Of the two options, paying content creators for what they produce seems much more "healthy" for society.
> or put 10x more ads on their page.
If the site becomes impossible to browse, I will leave.
> And those ads will frustrate you even more, because they aren't targeted to you.
I don't find targeted ads any less annoying than untargetted ones.
I like targeted ads, but I want them targeted to the content I'm consuming, not me personally. If I'm reading a blog or article on rock climbing, then, yeah, show me ads for all the gear. I'm interested. But don't show me an ad for an Instant Pot because I googled it last week. That's creepy.
That’s how Brave works—using zero-knowledge proofs, it uses your browsing history to inform its opt-in ad network on what ads should be shown to you. That info doesn’t leave your device.
Contextual targeting is happening, but it doesn't perform well as behavioral one yet while the risk is much higher for advertisers/publishers with the current technology (e.g. gun ads on a mass gun shooting news). With a sufficient technology, the situation can improve though.
The problem is they had a great solution with advertiser ID which is already used in apps. Expose that to websites and they would get a good compromise for both sides.
All consumer PCs except for Macs are subsidized by ads - the Intel Inside, Powered by Nvidia stickers and the crapware that comes on PCs.
Android phones are also subsidized by manufacture and carrier crapware. Not to mention that Android manufacturers get Google Play Services free in exchange for Google ads and analytics.
The most popular streaming media devices - Rokus - have hard coded buttons that go to the highest bidder. I have a Roku remote right now with an RDio button and there are some auth a Blockbuster button. Not to mention half the home screen on Roku devices are taken up with ads.
having to pay for quality content => bad (gee, who pays for NYT or WP after they're put behind a paywall?)
Getting the best ads possible without any human involvement => good
discovering the best products & services => good
microtargeting => bad
getting manipulated through political ads => bad
Look, ads ecosystem is the sponsor of free internet. Internet behind a paywall is very low utility. If anything, today's big ad tech are great democratizers & sponsors of the free web. That is, only if they can do the "good" and avoid the "bad".
> Look, ads ecosystem is the sponsor of free internet.
They are the sponsors of the shitty part of internet. They usually lead to the creation of crap articles and websites that use SEO to grab and resell traffic.
I think that even without ad revenue enough people are willing to put quality content online that we don't need to fear. There are more reasons that making ad money to create content.
Overall this doesn't make much of a long-term difference. Ad serving will shift to first-party with DNS or entire reverse proxies and is already in progress.
Apple has made some solid improvements but should now focus on not letting Safari become the next IE. The real impact changing the adtech industry is new regulations around data and privacy. That will change everything in a few months.
Good. If the ads are served from the first party domain that should make it a hell of a lot easier to attribute mal-action to that specific first party instead of offloading blame elsewhere.
Domains have nothing to do with the businesses and vendors behind them. The vast majority of publishers don't have the resources to run their own adtech infrastructure.
1st-party serving won't change anything with regards to chain of custody in ad content but the new regulations should at least make the chain shorter and more restricted in abilities.
> Domains have nothing to do with the businesses and vendors behind them.
TLS certificate says different. If malware comes from a TLS signed domain, then the owner of that domain is at-fault. Getting reparations from that domain is of course a different matter. But this should simplify the blame game.
> The vast majority of publishers don't have the resources to run their own adtech infrastructure.
I strongly disagree. Where there's a will, there's a way. And Google has the will.
No. TLS certificates have nothing to do with domain ownership or the businesses providing services behind them. Anyone can buy and transfer certificates. Besides there are actual business contracts in place so there's no need for such tenuous connections over domain names in the first place.
As for your disagreement, what's your basis for claiming that publishers can build, operate, and scale their own adtech infrastructure and salesforce?
> No. TLS certificates have nothing to do with domain ownership or the businesses providing services behind them.
> Anyone can buy and transfer certificates.
The registrar has a business relationship with the domain because the domain was purchased. The domain's owner uses DNS to point to an IP address. The client connects to that IP address and verifies that the IP address knows about the domain. "I paid for you to see example.com at 1.2.3.4" and "1.2.3.4 is responding to the name of example.com" completes that relationship.
With Letsencrypt, anyone can get certificates for free. And you certainly can transfer your private key and certificate information to a third party. That won't stop you from being liable for any malware served by them because you are the one who owns the domain. Unless you transfer the domain too. Then it's all on them. The third party. Hope you transferred the IP address payment too.
> Besides there are actual business contracts in place so there's no need for such tenuous connections over domain names in the first place.
Business contracts don't mean shit to me because I can't inspect them. I can inspect DNS, however.
> As for your disagreement, what's your basis for claiming that publishers can build, operate, and scale their own adtech infrastructure and salesforce?
Publishers can just as easily create advertisements and host them on their own website the same way they host any other static content. Scaling advertisements is just as easy (or hard) as scaling any other web services. Once upon a time businesses actually employed people for their sales departments instead of offloading it to a third party.
> "Business contracts don't mean shit to me because I can't inspect them. I can inspect DNS, however."
Your personal beliefs don't overrule how global business is conducted. Domain name ownership has nothing to do with business operations, liability or regulations. You can blame a certain domain all you want but it's ultimately meaningless.
> "Publishers can just as easily create advertisements and host them... Scaling advertisements is just as easy"
I have 15 years of experience in adtech and know 100s of publishers personally. It's not that easy and it's only getting more complicated as advertisers, campaigns and measurement get more sophisticated.
Please don't underestimate an entire 12-figure industry as if everyone else is clueless and you figured it all out. It would be no different then walking into Boeing and telling them how to design a plane or telling Dropbox you can code it all in a weekend.
> I have 15 years of experience in adtech and know 100s of publishers personally. It's not that easy and it's only getting more complicated as advertisers, campaigns and measurement get more sophisticated.
Advertising doesn't need to be sophisticated or complicated.
> Please don't underestimate an entire 12-figure industry as if everyone else is clueless and you figured it all out.
Please don't assume I don't understand how advertising works.
If the malware comes from a domain they own, it is their fault. If they can't be bothered to handle their own ads or pick a better ad vendor, they can pay up for the damage they served up to my computer.
Is apple offering an alternative for blogs to make money or should they have cake instead? Brave browser does it better than apple. I can already see webmasters sending users to download brave instead of safari
Apple doesn't care about the web (why should they?). Their iAds thing went nowhere.
(please enlighten me if apple is offering an alternative. i d love to use it)
There's another way to think about this. Privacy is a human right. If your business model is built on violating human rights at scale then I'm not going to have a lot of sympathy for you if technology kills your business.
The lesson from Apple is build something great, something that people love, and they will pay you for it.
They keep trying with Apple News & similar products, but frankly people aren't (for the most part) willing to pay for content they're used to getting for free. See the mobile app marketplace, e.g.
I really hoped Apple would get into micropayments for online content a long time ago, but I haven't even glanced at what Apple News+ entails, so perhaps I'm a hypocrite.
The plan should be to make something so good people will pay for it. Asking Apple to help with that is a little like asking a guitar company to help musicians make money.
if you go by that analogy, it's as if apple is taking away all the tip jars from musicians because they dont like them. wouldn't musicians be right to protest?
I would not hesitate to find a paid alternative to a free site/service that tries to control my choice of browser. The moment you try to push me into changing my up to date setup is the moment you’ve lost me as a user forever.
I'd like to see a p2p mesh network where articles stripped of ads are upvoted, annotated, and shared. Something like outline, but truly distributed. The markup doesn't allow for embedded javascript or cookies.
If it were to get popular, a microtransaction system could be built on top to pay the authors. In fact, payment could serve as one kind of ranking heuristic.
Store articles in a blockchain and garbage collect them from your node when they're no longer being actively read. Content will of course always be available as long as some node has the full history. This has the added benefit of making future updates or redaction without explicit change logs impossible.
You could even distribute comments the same way. All participants have keys, everything gets signed, natural interest graphs form...
This type of application of bkockchain tech seems more immediately useful than currency and brings us back into alignment with the old web paradigm. Distributed, experimental, hobbyist, not commercial.
Is a "garbage collected blockchain" even possible? There are retroactively modifiable ones, like git, but each git repository is its own world, so the trust issues are quite different...
there are plenty of systems to do that nowadays. it's a legal problem though, not a technological one. accounting and taxing those transactions is impossible under current laws
> In comparison, 79% of people who use Google’s Chrome browser allow advertisers to track their browsing habits on mobile devices through cookies.
Something to keep in mind the next time you see someone from Google complaining about how Apple won’t allow Google Chrome to run natively on the iPhone.