[manigandham]

This is a rather shallow article with barely any real points. Talking about all those externalities only serves to create noise rather than support any arguments against the ad model. You can just as easily argue that images and CSS should be removed to save even more bandwidth and processing power. Why even go online at all actually when you can get a newspaper for a few cents?

Here's the reality - We only have 2 options to enable the value exchange of content: direct (via you moving funds to the publisher) or in-direct (monetizing through ads). It's a binary choice. It's one or the other, that's it. There's no magic 3rd option that's available (today).

Here are a few reasons why advertising is still a better model:

- It's fast. Faster than money transfer.

- It's efficient. Money is not cheap to transfer, even today.

- It's passive. No mental energy expended or decision making required.

- It's anonymous. No need for payment details or 3rd party to handle funds.

- It's secure. Again no need for payment information, there's not much ads can do. (Malware is a different problem.)

- It's accessible. Your wealth doesn't determine the amount of content you can get. Everyone can access everything equally.

This is fundamentally a question of human behavior and the common reaction is to take things for free when possible. People just do not value content that highly, if at all. You can already see the reaction to rising rates for netflix and spotify even though they provide so much value, certainly more than an equally priced cup of coffee.

Add to that the amount of willpower and decisioning it requires to judge whether a piece of content is worth it or not and it's clear that this is not a scalable system. Perhaps a "cable subscription" for the internet with time-based billing for each domain might be something that works as a direct-payment option in the future but this is a massive technical, security and privacy challenge the way things are now.

I do agree that ads have gotten out of hand recently - however this is an implementation problem. Regulation and standards and enforcement are what's needed to fix this. It has nothing to do with the ad model itself which is still the most universally scalable and viable monetization system for content.

[Terr_]

Saying "it's just an implementation problem" understates the severity of that problem, and I think it's a little deceptive to claim that laws, regulations, and economic models are all "implementation" just as much as the raw code.

I mean, would you say that dumping a type of toxic waste into a river is an "implementation" problem when there's no law against it?

> It's fast.

Except when ad-frameworks slow down browsing to a crawl because they heap innumerable delays on the page-load in order to "auction" the spot and maximize revenue.

> It's anonymous

Except for our big ad-networks that incorporated quita lot of fingerprinting, profiling, and even surveillance.

> It's secure

Except for how ad-networks are a vector for malware, since there's no incentive to vet content beyond maintaining their own throughput.

[manigandham]

I believe you missed the statement where I said this is an implementation problem. I recently founded an adtech company to show how these issues can be fixed.

Ads are fast. Ad networks however are slow. Bad tech with old vendors is most of the problem. Our ads load faster than the content half the time. Sites like Washington Post, The Guardian, Ars Technica show how you can have a fast site with ads. Very solvable through technology.

Ads are definitely more anonymous than payments. Payments = credit cards or bank accounts. Guess what that means? Your name, address, age, credit history, employment, etc all perfectly accessible and tied to your identity. Profiling is used to give you more relevant ads. Someone who doesn't have children doesn't need to see diaper ads so it's a perfectly reasonable thing. I agree that sometimes it's taken too far with retargeting where you see the same ads constantly but this is a different matter. Our network focuses mostly on context of the page (tech site = tech ads) and performs just as well without any tracking (other than history of ads seen to avoid showing same ads).

The internet is a vector for malware. There absolutely is a way to vet content, but standards and enforcement is just not there right now. Ad networks dont really suffer anything when bad stuff happens so they have no incentive to do much about it. We don't even accept any 3rd party content on our network so we're far more secure than even the sites we run on.

Again, all of this is down to implementation - nothing wrong with the ad model itself.

[Terr_]

> I believe you missed the statement

I believe you posted a reply while I was still making postscript-editing. A bad habit of mine, I know.

[manigandham]

Got it. So with your edit:

> I mean, would you say that dumping a type of toxic waste into a river is an "implementation" problem when there's no law against it?

Yes. This doesn't mean we stop doing anything that produces toxic waste. Cars are also a vector for criminal activity. Like guns and knives and cellphones. We don't get rid of them either. Instead we use regulation and standards to enforce the safety mechanisms and proper ethical behavior.

[mirimir]

> Ads are definitely more anonymous than payments. Payments = credit cards or bank accounts. Guess what that means? Your name, address, age, credit history, employment, etc all perfectly accessible and tied to your identity.

So what does 15n1xkeDCGJJQAcYv17M4bPVfGqnksihLG tell you about me?

[manigandham]

Not much. Here's how our (modern) ad network works:

New user seen, gets assigned random UUID like ABC12345. Show an ad based on what we know about site context.

Show ad, track that ABC12345 has seen this ad. Next time, we check your history and know to show you a different ad. If you click on anything, we see that as a signal that you perhaps like that category of ads.

Eventually we have a list of ads you've seen (to keep the frequency of those ads in check) and what categories you seem interested in (based on your clicks). Combine this with the context of the site and we can give you better ads. We DO NOT magically get your name, address, birthday, or any other personally identifiable information.

--

Just for more detail: Facebook and Google are different. They are the biggest adtech players and own about 80% of the market not only because of their size but because of how much they know about you. Most people willingly tell both companies everything about their lives so they have a distinct advantage. They also know your identity across devices because of logins. These massive companies are a much much bigger privacy issue than any independent adtech company simply because independent players do not have access to all this detailed 1st party information (or it's provided through rough aggregated/statistical means).

[mirimir]

My point was that payments can be arbitrarily anonymous.

I'd prefer a micropayment system that was as automated as the current ad system is. Browsing a site would trigger an initial small payment. The total payment would be proportional to engagement, including the percentage of the article read, time spent on the site, links followed, and so on. There'd also be options to pay more, or to cancel payment.

I'm guessing that browsing would cost about $0.001, and that reading would cost between $.01 and $0.10 for average news articles, blog posts, etc. Magazine-length articles and reports would cost more. Is that about right to match typical ad income?

[manigandham]

Yea I mentioned this is my first post as time-based subscriptions would be perhaps the ideal instead of micropayments. The amounts you quoted are pretty fair.

The most scalable payments option is typical credit cards or bank transfers, which obviously come with identity information. Perhaps there can be an option for bitcoin but it's definitely not ready today. No company would deal with that at scale.

[kristianp]

That you know something about Bitcoin? :)

[mirimir]

I was maybe expecting some blockchain analysis ;)

[pdkl95]

Adding to what Terr_ already pointed out:

> It's efficient

Money may not be changing hands, but a lot of information is, which is easily worth more. Leaving out these costs are yet another reason why advertisers have a reputation for dishonesty.

Also, the article specifically discussed the terrible power efficency that is always going to be associated with the extra transfer and processing associated with fetching ads. Or do you think they get to violate conservation on energy to be downloaded? If you are adding ad to a page, you are making the client pay more to see it. Offloading costs like this - while pretending it is efficient - should really be criminal.

> It's passive

Bullshit, and you know it.

The very idea of advertising is to be distracting. The arms-race we've seen over the last few decades where advertising went from simple banners, to popups, to various other dynamic rendering tricks, to the modern modal/click-through video ads - and worse - are all about not being passive. This is suck a basic concept in advertising that I have to conclude you are dissembling deliberately, or are such a "true believer" in advertising you cannot see it any other way.

Also, it takes an incredible amount of energy to ignore ads. Anybody that thinks ads don't cost any mental energy is either lying or addicted. Install an adblocker and use it 100% reliably for several months to break the addiction, and you will be shocked at the energy advertising wastes when you see it again.

> It's accessible

Except if someone is poor, they cannot afford whatever it is you are advertising. I've seen this meme a lot recently, so it's obviously a talking point the advertising indujstry is using to try to justify their drain on society. You're not fooling anybody, and this kind of appeal only makes advertisers look desperate.

--

While Terr_ already mentioned this one...

> It's secure.

You don't get to exclude malware, when advertising is the vector.

Even ignoring that, leaking tracking information makes you malware!

[manigandham]

>>> Also, the article specifically discussed the terrible power efficency that is always going to be associated with the extra transfer and processing associated with fetching ads. Or do you think they get to violate conservation on energy to be downloaded? If you are adding ad to a page, you are making the client pay more to see it. Offloading costs like this - while pretending it is efficient - should really be criminal.

These are externalities. Do you get upset that articles have images in them? What about products coming in packaging that you just throw away? Isn't that gas and energy you are wasting on something you don't want?

Power efficiency of ads is really not that big of a deal.

>>> Bullshit, and you know it.

I was talking about the effort involved in choosing to pay for each pageview. Not ad formats. I believe you're conflating the two together.

>>> The very idea of advertising is to be distracting.

Actually the idea is to gain your attention. And yes, we already know this because it's quite literally what advertisers are paying for. Your attention.

>>> advertising went from simple banners, to popups, to various other dynamic rendering tricks

Yes, this is an implementation issue where ads have gotten out of hand. I agree with you that we need to return to better standards. However this doesn't mean the advertising model is broken. It's still magnitudes less effort required than having to decide to pay everytime you load a new page.

>>> to ignore ads

The point isn't to ignore ads. Advertisers are paying for your attention which pays for the content. It obviously wouldn't be a very good model if they never got the attention that they paid for. Yes it takes energy to ignore, also takes energy to choose to pay. Ads are generally easier to deal with.

>>> when advertising is the vector

Everything is a vector. Pick any object in the world. It's a vector for use by a criminal. We should focus on stopping the criminals, not banning everything.

>>> leaking tracking information makes you malware

Genuinely don't understand this. What tracking information and what is harming you?

[pdkl95]

> Power efficiency of ads is really not that big of a deal.

Saying that doesn't make it true. The article says otherwise, and I believe it. I really don''t understand how you think this could be true.

Also "not that big of a deal"? So you think it's fine if I quietly tap into your power lines and steal some of your electricity, as long as it's only a small amount that "isn't that big of a deal"?

> Do you get upset that articles have images in them?

Do you have a hard time understanding the differenced between the content I requested, and crap added in that has nothing to do with the content I requested, which is specifically designed to distract me from that content?

If you truly cannot see the difference here, I don't know how to help you.

> I believe you're conflating the what I mentioned about the effort involved in choosing to pay vs types of annoying ad formats.

NO, I am NOT.

The type of ad format isn't particularly relevant. Advertising - any type - is by definition designed to distract people that see them, and it takes a significant mental effort to work around these distractions.

Be aware that if you want to argue against this point, you're challenging one of the basic laws of marketing as it has been practiced for many decades.

> Actually the idea is to gain your attention.

Thank you for admitting advertising is trying to be distracting. That is, the opposite of "passive". You seem to be arguing against your previous claims.

> having to decide to pay every time you load a new page.

This is a complete straw-man argument. You're suggesting that "paying every time" is the only alternative to advertising, which is patent nonsense. Even in the case of a paywall model, nobody would "pay every time"; subscription models have always existed.

Are you going to pretend that Patreon doesn't exist? Many models of funding exist and are being used successfully.

> Everything is a vector.

If you don't understand the concept of minimizing the attack surface, you really should refrain from making security claims. Advertising is adding additional attack vectors.

Besides, it is incredibly dishonest to try to say advertising is on equal footing here. Ad networks have been the vector of choice for a long time. Some of that is because of their utter disregard for security. They are also chosen as a vector because ad networks by definition reach a wide audience.

> What tracking information and what is harming you?

I'm not sure how to respond to this. Are you seriously ignorant of the tracking done by ad networks? Or did you miss the last few years of data breeches that expose information and the powerful new machine learning tools we have to analyze large databases?

If you are gathering information - ANY information - about what people do on the internet, you are creating a target. If you are aggregating that information - again, ANY information - across sites, even if you supposedly anonymize[1] it, you are creating datasets that are the most tempting target imaginable for insurance companies, extortionists, stalkers, and governments with national security letters.

In a sane world, you would be liable for any damage caused by that database you created.

[1] As DJB says[2], "Hashing is Magic Crypto Pixie Dust, which takes personally identifiable information, and makes it incomprehensible to the marketing department."

[2] https://projectbullrun.org/surveillance/2015/video-2015.html...

[manigandham]

I honestly believe you're either taking this personally and/or using emotional arguments from some deep hatred of advertising. Nevertheless I hope this clears it up for you:

---

Advertising is meant to gain your attention. This is not a secret conspiracy, it is quite literally the entire business. That being said, there is vast differential in the amount of energy expended in ignoring content vs the amount needed to make a willful decision of how to value and purchase said content.

Subscriptions are actually even harder to decide on because they are a much higher cost that requires a commitment with a single outlet and trust that their quality will continue to be to your liking vs a single article.

Patreon is not an alternative funding model - it is still direct payment/funds transfer. It's just optional. It's also not viable for any scalable business. Not everything can or should be a small little site.

Advertising is packaged with content because the publisher owns the right to that content and how it's distributed. This is the same in every other medium ever invented. Same with any other product you buy. You get it as it comes. If you're not compensating that advertiser for that content directly, then they have every right to include advertising to offset the cost of creating and distributing that content to you which you requested. Would you rather they deny your request?

Everything is a security risk. You should stay in a nuclear bunker and off the internet to truly minimize surface area. However that's not realistic. Ad networks definitely have an implementation problem with not enough standards on malware testing, and yes they're a good choice because of their scale. However guns are also efficient at killing and cars are efficient at escaping, should we also get rid of both because they're used by criminals?

Any database containing data and thus value becomes a target. Should we also stop creating any kind of database because it might become a potential target? So medical records, court proceedings, tax records, insurance claims and every other database should immediately be destroyed because someone might want to access it? Again the real problem here is the fact that criminals will be criminals, not the tools they use or targets they go after. There should absolutely be more safeguards with ad networks along with liability as I've stated in previous posts (and I'm doing my best to make that happen) but arguing that we should get rid of the advertising model because of some security issues does not make sense.

[pdkl95]

If I was taking this personally, this reply would simply be a flame (or I wouldn't have bothered). I am writing this reply for the same reason I write most of my posts here on HN: education, and the hope that the tech industry can be redirected away from the problems the industry has cultivated in the last decade(s).

> This is not a secret conspiracy.

I never said anything close to that, and I'm not sure whee you got that idea from.

> Not everything can or should be a small little site.

Obviously. Many things can be done small scale, and some people even prefer that for a varietyh of reasons. Patronage is an alternative funding model. It should go without saying that there isn't a 1-to-1 mapping between possible funding models and the projects that want funding. Clearly patronage will only work for some situations.

> they have every right to include advertising to offset the cost of creating and distributing that content to you which you requested

Yes, they do.

Just like the end-user to which they gave a copy of their content has every right to view that advertising, or cut it out, or pay someone to cut it out, or simply decide they want to go somewhere else. Creative works have a copyright, which does not extend to use.

Advertising has always involved risk.

> security

If you want to actually learn the problems of the digital exhaust (or data trail) we leave behind, I suggest listening to some of Dan Geer's talks in the last 1-2 years. The only reason this isn't a problem right now is that the tech is still new. However, the last year of leaks and compromised databases has shown that the peace will not last.

> criminals will be criminals

Yes. Exactly. Which is why it is important to not let power concentrate, and the databases that advertisers are building (with Google and Facebook at the top) are a power that is new to the human experience. The solution is to limit how much damage an be done by limiting how much data can be aggregated.

You might characterize my goals here as a type of self regulation, as I fear the blowback as people finally figure out what technology has been doing. The alternative to self regulation is some type of Butlerian Jihad (probably misguided and poorly targeted). This isn't really hyperbole as these are two extremes and reality is usually somewhere in the middle. That said, when I talk to non-technical people about how modern tech works (adsense, other tracking by ad networks/etc, "smart" TVs, and other supposedly-"free" services), there is a common first reaction: "Why aren't they in jail?"

> guns

Are a bad comparison in many ways. The risk from a criminal with a gun stops after they are caught. The entire problem with data is that it is a permanent risk.

> (and I'm doing my best to make that happen)

For that, my applause and thanks!

> arguing that we should get rid of the advertising model

That is not my goal! Your replies are often arguing with straw men.

I believed advertising is on shaky ground now that General Purpose Computer has given powerful tools let people view data as they desire. Adblocking is only the beginning. This is a pragmatic statement, because I've been watching a backlash against advertising happen for several years now. However, advertising will never go away entirely, as it is one of the useful funding models.

My point in arguing is a refutation of the well-known fallacies and talking points in your initial post. You have argued against some (that advertising is in any way passive - as you say, it is "meant to gain your attention"). You have also made some patently false claims about power and anonymity (HTTP requests are generally not anonymous, even without the crazy tracking that the ad networks do). You also have made numerous claims that elevate advertising as the only or majority option for funding something on the network. I've already mentioned security above.

These specific claims are what I have a serious problem with. Quite a few funding models exist, and many variations are being explored. This is one of the huge benefits of modern technology: it made available so many more options in every area, including business models. The human experience with internetworking is still in its infancy, and I suspect we haven't even discovered the best ways of utilizing the network for business purposes.

[sarciszewski]

> - It's secure. Again no need for payment information, there's not much ads can do. (Malware is a different problem.)

http://www.wired.com/insights/2014/11/malvertising-is-cyberc...

Advertising is an attack vector. A security risk for the end user.

[manigandham]

The entire internet is an attack vector. Malware has existed for a long time in numerous forms, ad networks are just another possibility but so are websites themselves, downloadable software, mobile apps, etc.

As mentioned in another comment - We also have cars and cellphones that can be used by criminals. That doesn't mean modern transportation and communication has a problem, it means that criminals will be criminals and we need better ways of dealing with them. As far as I can see, that's an implementation problem.

Here's the truth: I've been in the ad industry for a long time and there is literally 0 enforcement of anything. Lots of fraud and shady practices but no regulation and nobody to answer to. If you do get caught, change the name of the company and try again. I can guarantee that if we actually send people to jail for not vetting malware or running illegal or shady ads, this problem will pretty much go away immediately.

[panic]

You could do indirect payments without ads. For example, say an organization gave grants to ad-free websites based on certain criteria. Some of these criteria could be how much people visit the site or how highly they rank it. Funding could come from people interested in keeping the web ad-free.

[manigandham]

This is still direct via donations. Just by a few doing it for the many.

So the question is, who are these special few?

NYT times costs hundreds of millions to run. That's just a single publisher. Most major sites require somewhere in the 8 figures. As theory and data have shown, human behavior is not conducive to paying if you can avoid it. Unless your plan is to turn content publishers into tax-funded state-run companies, I fail to see how this could possibly work.

* Before it's mentioned: Wikipedia is perhaps the only example of donations working at scale. However, it's not really working because Wikipedia doesn't produce any content. It's also not a business. Wikimedia which is an actual business runs the wikia.com network of sites and they're all monetized through ads.

[pdkl95]

> NYT times costs hundreds of millions to run.

Maybe their product isn't worth hundreds of millions of dollars.

As for willingness to pay, you seem to be projecting your own selfish behavior onto others. I know quite a few counter-examples that prove people are willing to pay for content they actually value. They have embraced technology to lower costs, and have cultivated very generous audiences using places like Patron.

> only example of donations working at scale

LOL. Only if you don't look. Also, why do you think everything has to work "at scale"? Small-but-loyal audiences are fine.

[manigandham]

> As for willingness to pay, you seem to be projecting your own selfish behavior onto others.

I believe this is what you're doing with your comments. A few counter examples are great but we already have billions of people online constantly generating data and plenty of experiments that have shown that donations do not work.

> Also, why do you think everything has to work "at scale"?

Because the internet is at scale. I'm not talking about a specific site but rather the donation model in general. Very few actually donate and thus it cannot sustain the quality and quantity of content that is available today. A few site/podcasts/shows/whatever might be able to work with this but it will not work for the internet as we know it today.

Value is subjective, just because it might not be worth direct payment doesn't mean it's worthless (ie: Facebook). As a common example take a look at adblocking: if you don't like the ads on a site, then you should stop visiting the site. By using adblocker and continuing to go to the site, your actions show that you find value in the content but do not care to let the publisher be compensated. This is just one of many manifestations of the human behavior I've described where donations by the few does not support content for the many.

[panic]

If enough people care about an ad-free web, it wouldn't have to be a special few: it could be a relatively large group of passionate people.

But that's just one idea. The point is that it's not a dichotomy of low-friction, indirect ads versus high-friction, direct payment. There may be other options.