Hacker News Reader

The problem isn't advertisements per se, the problem is advertisement networks that track users and sell their profiles. The only thing I have installed that counts as an ad blocker is Ghostery. I still get served advertisements at conscientious websites like duckduckgo.com or any other place that doesn't rely 100% on tracking networks.

Imagine if simply because you walked into Target, they hired a private investigator to follow you around and determine your personal habits, hobbies and other stores you visited. Imagine if they then used that information to attempt to lure you into the store, or sold that information to Best Buy so they could lure you into the store. I think most people would have a problem with such behavior. Making a more realistic physical analogy would be slipping an RFID token into your wallet without you noticing and only tracking you in stores that use the corresponding RFID reader.

In my opinion, tracking networks go the extra mile beyond a creepiness factor. My visit to your website isn't tacit approval for you to peruse my browsing of other websites. I shouldn't need to opt out of this behavior by installing Ghostery in the first place.

Individual profiles are rarely sold directly[1], but some demographic is taken so that the publisher can sell your traffic at a higher rate. This is a good thing because it means they can make more money with less traffic, which in turn means that they don't have to appeal to anybody and everybody. This is (hypothetically) where quality content comes from.

Unfortunately most of the ad industry is really crap at this.

For example, Oracle/Bluekai leak `var bk_results` into the web page allowing anyone to pick up this data which means that this information can (and is) often used for much more than just better ads.

[1]: One notable space where they are sold is ABM. Unless you're a decision maker for an enterprise supply budget, this isn't you.

It is odd to see advertisers speak of tracking so casually. Tracking is still new. Up until a few years ago (like 3) the majority of the industry (by money) did not track users. A huge, but admittedly declining, segment still don't.

Billboards do not track. TV ads do not track. Physical shop fronts do not track. Radio ads do not track. And a great many website still sell space to people selling actual products, rather than banner ads, which bypasses all adblockers.

There are plenty of ways to get your message out, and make money in return, without inventing new supercookies for me to ferret out of my system.

Both Outdoors and Television do the same by-demographic buying/selling, in fact, that's where the Internet got it! Outdoors and Television don't track (except when they do[1]) because they don't need to; because there exists a huge problem with digital that doesn't exist for Outdoors and Television: Ad Fraud.

There are a lot of websites out there with a lot of (aggregate) traffic, but that individually don't have enough traffic to approach advertisers directly. It is good that we can monetise them because sometimes this can fund their niche content, however it also means that people can make fake websites and trick advertisers into buying that space.

Of course, real users don't go to those sites, which is a nascent use for tracking: Sites that are visited by some population who don't visit other sites are simply more suspicious.

[1]: Precision Marketing (for example) has data on billboards collected from mobile devices. Many OTT and SmartTV devices send tracking data as well.

Have there been studies showing how much, if at all, user tracking improves ad clickthrough or revenue?

Yes to both, but I am suspicious about most of the mathematics used to prove uplift and conversions using micro-models because of simple counter-examples like the blank ad[1] that generated a median clickthrough rate, and because of the prevalence of one-tailed "dropoff" charts don't pass the laugh test.

For more on this phenomenon, there's a book[2] that I'm oft to recommend.

Nevertheless, I also believe there are genuine means for revenue that utilise user tracking, but this is more work and attention than the typical digital ad campaign receives.

[1]: http://adage.com/article/digital/incredible-click-rate/23623...

[2]: http://www.amazon.co.uk/How-Lie-Statistics-Penguin-Business/...

> TV ads do not track.

Are we sure about that? Smart TVs would seem to make this at least possible, and probably easy.

Of course a smart TV tracks you. That's why the manufacturer made it (to earn money from you over its lifetime). The other features it has are just the bait on the hook.

Which is why I tell everyone to avoid SmartTVs and instead use some form of media player to support networking. It means yet another remote, but given that my TV remotes have already been usurped by my isp-provided decoder box. (Also a good idea as players are cheaper to replace than TVs.)

TVs may not track yet, but phones or PCs are already tracking the TV ads people are watching:

"Beware of ads that use inaudible sound to link your phone, TV, tablet, and PC. Privacy advocates warn feds about surreptitious cross-device tracking."

http://arstechnica.com/tech-policy/2015/11/beware-of-ads-tha...

TV ads can track using inaudible watermarks picked up by other household devices (especially apps running on phones; arstechnica wrote about this). Probably similar with radio ads.

Physical stores track using Wifi, Bluetooth, and cellphone signals. Billboards could conceivably do the same.

The world is creepier than most consumers imagine, and it needs to stop.

>> "Billboards do not track."

A competently-designed out of home ad setup will probably at least make an attempt to track its audience, often by doing some sort of telemetry to your smartphone or by literally photographing you as you pass by. [0-1] I don't work for these outfits but a lot of people erroneously believe the "billboards don't track" claim.

[0] http://web.admobilize.com/ [1] http://www.quividi.com/applications.html

Do you ever feel that companies that do online advertising are not the best use of your talents? Is there no choice in life? Nothing better to work on?

I own a small, independent mobile games company. If you can figure out how to get people to pay for games (without resorting to psychological trickery like Candy Crush) you could make millions.

I'd love to run a pure paid model and rip out the bloat and headaches of advertising SDKs, but I also have employees who count on a paycheck and like working on games, and app stores that foster a race to the bottom and highly promote free games over paid.

So I'm pretty happy with my place in life right now.

Keep chasing that dragon. Mobile games aren't much more than ringtones. Take note of how much time you spend marketing and whale-chasing v how much time is spent on the actual game.

KSP, Subnautica, Prison Architect, Minecraft, Don't Starve ... there is plenty of money being made by independent studios once you get off the mobile merry-go-round.

This is incredibly, offensively tone deaf. Our biggest game is over 5 years old, we spend 95% of our time on it on features and improvements, and we don't have in-app purchases (besides the ability to buy away ads) so we could care less about whale chasing. Millions of people play, our retention is at the top of the market, and we make 95% of our money advertising because less than 2% of players are willing to part with $1.99 to turn off ads (but many play for years).

Please don't make assumptions or comments about things you don't actually have experience with.

Lol, no experience. I've lived through the mobile gaming thing. I've seen clients rise and fall, and cry, because of the ridiculousness that is selling games on phones. It is dragon-chasing. It's all about trying to pry pennies out of people for tiny little games. 1.99? Right there is another silly marketing trope, as if that penny would ever make a difference for a careful consumer.

What game is this. Curiosity compels me to see if it's worth 2 bucks to turn off ads.

Probably http://flyclops.com/games/domino.html (looked through his HN description)

I've been working on news sites for 7 years. I rarely comment here, but that's a good question.

1. No, I think the opposite. One of the reason that Internet advertising has gotten to this point _because_ smart people don't want to work on it.

2. The history of mass media is tied to the history of advertising. I know its trendy to say things like "the market will innovate" followed by no new ideas. This isn't the answer. The answer is to fix web advertising's problems.

3. I love the writing. I think it matters. I meet people who read my sites all the time and they agree. The survival of this depends on solving a series of really difficult, interesting problems.

Maybe it's just me, but I think that's cool.

Re 2., of course "the market will innovate" is bullshit. The market is perfectly fine with the status quo. Moreover, the market is a system that optimizes for efficiency of making profits and only for this. All those complains about ads are in fact complains about the market being too efficient at extracting profit from users. If we want to use the power of the market to fix abusive ads, we need to destroy their profitability.

"... smart people don't want to work on it."

That's encouraging.

At the same time, there are some clearly talented authors of networking code working in advertising (well, at least one). Impressive.

I cannot comment on the idea of "markets" but I can offer an opinion about people. If they can work on what they _want_ to rather than what they believe they _are forced_ to, I think that's cool.

Dennis Ritchie once said in an interview he could have chosen to work on calculating missile trajectories or some similar "difficult, interesting" problem, but instead he chose not to, and ending up working something like C.

I'm not sure anyone asked him to volunteer that information, he just did it. If my curiousity oversteps a line, I apologize. I wish I had the programming talent that I see some people (few) have and I just wonder what it must feel like to have that talent.

I noticed you get a 256-trie with:

    T:{$[~#y;z;@[256#x;*y;:;.z.s[x@*y;1_y;z]]]}

that t:T[t;"key";leaf] can be used to store a leaf in the trie, and t@/"key" can be used to look up the key in the trie.

Unfortunately, at least on my 1.7Ghz i7, with 8-byte keys, bin remains faster until around 1m records, and the difference remains negligible until you have more like 10m records.

By all means, get in touch. I'm always interested in hearing about better things to work on.

The problem isn't advertisements per se, the problem is advertisement networks that track users and sell their profiles.

I'm not sure what is gained by changing the subject like this, but I think at the end of the day we can agree that websites are allowing the ad networks they use to have nearly-unfettered access to your browser, and that this is a problem among several.

Why should the advertisement networks be the sole responsible when it is the publisher that subcontract the activity of delivering ads to visitors? The responsibility for distributing malware should be shared between each party that earn revenue from the illegal activity, which include the publisher, the network and the criminal. Passively earning money on criminal activity should not be the primary method that content creator earn money.

I also find it very problematic that offline advertisement are held to a much higher standard and required to follow local laws, while web based advertisement can even get away with distributing malware without repercussions.

Who said anything about passively earning money from illegal activity?

I'll clarify - by "we should hold networks responsible" I mean "we the publishers" - that doesn't absolve our responsibility to the consumer. The fact is, the overwhelming number of ads online are free from malware, to a large number of 9s-decimal places. Still, networks need to vet the ads they serve. That's something they _can_ do that publishers cannot.

The publisher has the same ability to vet the advertisement that a subcontractor delivery, that a company has to vet the construction material when subcontracting the construction of a building. If publisher were held responsible, they would only deal with advertisement network that are held economically responsible if advertisement they deliver breaks the law. That mean higher costs, less revenue, but an end to the wild west we currently have.

If I buy a paper based newspaper, the ads on those follow a very strict set of local laws. Each time an illegal ad is mistakenly printed, its news that other papers just love to write about. If I buy a online subscription, none of this laws are in place, and to the same large number of 9s-decimal places that you talk about, the illegal ads will be on the online version.

This is patently false. We have zero ability to vet ads as a publisher without spinning up an entire as sales department larger than my entire company. The networks we use can and do vet the ads before serving them.

Of course, you're right. It's actually very simple. You as a publisher need to verify ad networks. Ad networks should verify their ads. It's your responsibility as a publisher to find ones that do it properly. Because I, as an user, don't care about the byzantine world of advertising. I deal only with you, and if you serve me crap, I don't care if it's your subcontractor's fault. It's an implementation detail. I blame you, because I interface with you.

And you are absolutely right. We can't pass along responsibility and feign ignorance. It did take a while for us to do it right (whittle down the networks we use to the best and most responsible) and that did take some relationship building - but I'm sure glad we did.

  The networks we use can and do vet
  the ads before serving them.

You know this for every ad for each network... exactly how?

If they failed in this duty, would you be informed?

Do they have a duty-to-defend clause in your agreement with them such that if you are sued for serving malware in your app, the network will pay all costs for your defense and all necessary restitution?

There's a sour "we were only following orders" flavor to such an indifferent stance.

Right, but why do you stop there and blame Forbes and not e.g. Google -- the company that let the attacker upload the malware?

It's very hard for a publisher -- even one as big as Forbes to hold Google accountable for anything. The fact that most people have such a fundamental lack of understanding about how digital advertising works doesn't help either.

As a casual browser I can only blame the site that "served" me, good and hard. I have no idea what network or subnetwork ultimately served the malware/ad.

And since the visited site is the one that came into my "home," it's entirely appropriate to hold the site's feet to the fire.

When a site becomes well known for holding itself responsible for what comes out of the site, and doing it effectively, I'll whitelist them.

The reality is in fact very simple. Ultimately, the buck stops with Forbes. They're the ones who decided to serve ads, they've made the decision to do it without their oversight - so they're the ones to be held responsible if the ads start spreading malware.

It's publishers, not us, who can and should exert market pressure on ad networks to change.

Ultimately the buck actually stops with the end user. They're the ones who decided to ask a server outside of their control for any and all data it has stored at an address and then used it in a way harmed themselves. It very much is us who can and should exert market pressure on ad networks to change. The publisher, ad network, ad creator, any and all business, only exists to satisfy our needs and wants. We are the market not publishers.

Oh, but we do. We install ad blockers. That's the best way to exert that pressure from our side :).

> The publisher, ad network, ad creator, any and all business, only exists to satisfy our needs and wants. We are the market not publishers.

You'd wish. We are as much a market as dogs are the market for dog food. They're not, dogs don't have money. Dog food is marketed to human owners. And so it's publishers who are customers of ad networks.

>Oh, but we do. We install ad blockers. That's the best way to exert that pressure from our side :).

That's exactly what I meant haha. I find the constant barrage of news about how terrible ads are baffling. There's a very simple solution for end users and it's been around for a long time.

>They're not, dogs don't have money.

We do have money. That's what they want from their ads, our money. Don't view ads -> don't click ads ->don't spend money on advertised products -> problem solved.

Why do you think Google was the ad network responsible? The article names several other ad networks (Atomx, Yahoo) and doesn't even mention Google.

"Holding networks responsible" sounds great, but the money chain isn't transparent, so whether you hold them responsible or not is irrelevant: Users hold you responsible if they see your URL and have a bad day.

By the way, your website is down.

That's what I said - Forbes needs to hold the networks responsible. Obviously the consumer can't, and as I said in the edit, that doesn't absolve Forbes.

Also I don't spend any time on my personal website these days. I haven't checked in on it in months...

I wasn't disagreeing with you: Forbes absolutely should hold the networks responsible, because we hold Forbes responsible. However, the consumer can hold the networks responsible: they just install an ad blocker. :)

But I don't get it?

If i go to forbes.com and my computer is p0wned by some shit that they or their partners delivered, why are they not liable for damages?

Isn't this a class action suit waiting to happen, or is it not applicable, because Internet?

  why are they not liable for damages?

The more relevant question is: how would you (or another random end user) prove that the malware was served through their, and only their, site?

There's also a subtle problem of mistrust in all this.

Ad creatives have evolved from just simple banner images to more rich media but those files are relatively safe. The biggest problem is the explosion of javascript tags used for verification and analysis because a lack of trust and politics has led agencies/advertisers to require layers of external tracking.

This requirement has forced ad networks to support 3rd party javascript tags - which is basically impossible to verify for security. If you can run javascript on a page, you can control that page.

Fundamentally there needs to be more regulation, standards and actual enforcement to fix this.