What Happens Next Will Amaze You • Hacker News Board

Re: What Happens Next Will Amaze You

#111

Post by jonstokes » Wed, Sep 23, 2015, 2:05 PM UTC

This is a great piece, but I have to post a correction for this: "It boils down to this: fake websites serving real ads to fake traffic for real money."

There are, in fact, also real websites serving real ads to fake traffic for real money. When publishers promise an advertiser impressions or downloads or some other metric on a campaign, and then the traffic fails to materialize, the marketing department can always call in the bots. It happens at some of the biggest and most reputable sites out there, and what's crazy is the agencies (middlemen who buy ads on behalf of publishers) know it and don't care because hey, the metrics got met and they got paid their cut!

Re: What Happens Next Will Amaze You

#112

Post by AlexandrB » Wed, Sep 23, 2015, 2:05 PM UTC

Retra wrote: post #102

mozumder wrote: post #2

The correct answer is to never use programmatic ads. Ever. If you own a site, just curate a set of ads on your site that actually enhances your sites appeal. A fashion site should have matching fashion ads. A tech site should have matching tech ads. A local site should have matching local ads. Reach out to advertisers to get these ads. Never use an algorithm. You, as an editor, should be better than the algorithms. A…

Why would Vogue show you an SSD ad unless they've got data suggesting you want to see it?

Why would Vogue show an SSD ad at all? What does Vogue know about SSDs? Nothing. Why would I think that an ad for SSDs on Vogue tells me anything about the quality of the SSD they're hocking? It's non-information and communicates nothing about the product beyond the fact that it exists. Anyone who knows what an SSD is knows where they can buy an SSD - what I really want to know is which SSD should I buy and an SSD ad on Vogue doesn't give me any information in that regard.

Re: What Happens Next Will Amaze You

#113

Post by Retra » Wed, Sep 23, 2015, 2:08 PM UTC

AlexandrB wrote: post #112

Retra wrote: post #102

Earlier quoted context omitted.

Why would Vogue show you an SSD ad unless they've got data suggesting you want to see it?

Why would Vogue show an SSD ad at all? What does Vogue know about SSDs? Nothing. Why would I think that an ad for SSDs on Vogue tells me anything about the quality of the SSD they're hocking? It's non-information and communicates nothing about the product beyond the fact that it exists. Anyone who knows what an SSD is knows where they can buy an SSD - what I really want to know is which SSD should I buy and an SSD ad…

Your complaints have nothing to do with Vogue, but with the uselessness of an ad in general. Even if a site were talking about SSDs, you wouldn't need an ad to tell you about them, would you?

Re: What Happens Next Will Amaze You

#114

Post by spdustin » Wed, Sep 23, 2015, 2:11 PM UTC

benihana wrote: post #84

mozumder wrote: post #2

The correct answer is to never use programmatic ads. Ever. If you own a site, just curate a set of ads on your site that actually enhances your sites appeal. A fashion site should have matching fashion ads. A tech site should have matching tech ads. A local site should have matching local ads. Reach out to advertisers to get these ads. Never use an algorithm. You, as an editor, should be better than the algorithms. A…

> The correct answer is to never Ah sweet, it's that one commenter who can prescribe the absolute correct answer to very complex situations with a couple of simple platitudes! Since you seem to be in a mood to drop the correct answer, maybe you can help me with some difficult decisions I'm facing: what programming language should I use? what should the next move of my business be? is she really the one, or should I m…

I won't speak to the sarcasm-as-argument response, but I will say that there is a sentiment behind it that I do agree with.

I say - with differing commentary after the first 4-5 sentences - in every class I teach:

"I don't like the phrase 'best practice'. I won't use it. It's presumptuous and condescending. What's best for me might suck for you. Who am I to know your business requirements? There is only one true 'best practice', in my view, and that is simply to have practices that you and your team follow that serve to further your business goals and meet your business requirements. SharePoint isn't a best practice, but a system to support a corporate culture of collaboration or information sharing is a practice that many businesses can agree with. jQuery isn't a best practice, but using a framework that improves your team's productivity and helps it to meet your business' requirements - that's a good practice to follow."

We need a new phrase that gets the point across without the cognitive burden of being condescending and presumptuous. Consensus practice? Common practice? Popular practice? I don't know. But I know that "best" is a superlative that is too often used to disguise a marketing goal or to dazzle the listener/reader into accepting the superiority of the person using it.

Re: What Happens Next Will Amaze You

#115

Post by seiji » Wed, Sep 23, 2015, 2:11 PM UTC

vidarh wrote: post #107

seiji wrote: post #78

Earlier quoted context omitted.

adtech is a modern example of the banality of evil: http://opinionator.blogs.nytimes.com/2013/09/15/the-banality... . It's not secret Machiavellian billionaires running around shoving ad tracking upon us, it's ourselves just trying to get a tiny glint of social approval for out thinking everybody else without regard for consequences at scale.

And it's so easy. At a past company we once decided we'd test some assumptions about one of our forms by capturing keyboard events. As in, we added a handler for keypresses that'd call back to the server, so we could capture and play back exactly how users interacted with the form. It provided some useful insights (one thing that surprised me was how bad people were at recognising and operating drop downs), but we ve…

Or they might write something, change their mind and enter something else

Facebook explicitly does that (log every field input even if never "submitted") and uses unsubmitted thoughts to build more detailed models of you.

The "trace-per-character" behavior terrified me when Google Suggest first launched. It's a neat tech example, but there's no way people can understand "everything I type, live, as I type, will be recorded forever." Then they made it default Chrome bar and default google search behavior. Average users (heck, even technical users) have little idea how much data internet services capture about them though non-explicit interactions.

Re: What Happens Next Will Amaze You

#116

Post by estefan » Wed, Sep 23, 2015, 2:15 PM UTC

AlexandrB wrote: post #112

Retra wrote: post #102

Earlier quoted context omitted.

Why would Vogue show you an SSD ad unless they've got data suggesting you want to see it?

Why would Vogue show an SSD ad at all? What does Vogue know about SSDs? Nothing. Why would I think that an ad for SSDs on Vogue tells me anything about the quality of the SSD they're hocking? It's non-information and communicates nothing about the product beyond the fact that it exists. Anyone who knows what an SSD is knows where they can buy an SSD - what I really want to know is which SSD should I buy and an SSD ad…

> what I really want to know is which SSD should I buy

In that case you need reviews, not ads.

Re: What Happens Next Will Amaze You

#117

Post by idlewords » Wed, Sep 23, 2015, 2:19 PM UTC

jonstokes wrote: post #111

This is a great piece, but I have to post a correction for this: "It boils down to this: fake websites serving real ads to fake traffic for real money." There are, in fact, also real websites serving real ads to fake traffic for real money. When publishers promise an advertiser impressions or downloads or some other metric on a campaign, and then the traffic fails to materialize, the marketing department can always c…

Thank you, I appreciate the correction! That tactic hadn't occurred to me.

Re: What Happens Next Will Amaze You

#118

Post by spdustin » Wed, Sep 23, 2015, 2:19 PM UTC

I suspect it won't be long before ad networks supply self-hosted proxies in multiple languages for sites owners to use. Or, since ad blockers primarily use DNS names, ad networks could simply add CNAME support to their infrastructure. Granted, either approach would injure the ad networks' ability to track you, and expose those cookies to the site owner, but I would suggest that fingerprinting tech is mature enough that proxying the request won't prevent the ad networks from correlating identities.

It's cat-and-mouse. We're enjoying the benefits of a faster, safer and cheaper web (as are the developers behind those content sites using the ads, I reckon), but my prediction is that those benefits will be short lived. There are cleverer people than the browsing masses, they will find a way.

Re: What Happens Next Will Amaze You

#119

Post by jonstokes » Wed, Sep 23, 2015, 2:23 PM UTC

Ok, this is a fantastic piece that says a lot of useful and thoughtful things, some of them critical about the priorities of the SF startup scene. And what are a nontrivial portion of the HN comments doing? Complaining that he was unfair to Elon freaking Musk!!

This is exactly the sort of techno-douchebaggery that the author is writing about, so way to prove his point, nimrods.

The one redeeming quality of such comments is that they fall into the category of "HN comments that are best read in Comic Book Guy voice", and when read that way they're good for a chuckle.

Re: What Happens Next Will Amaze You

#120

Post by coldtea » Wed, Sep 23, 2015, 2:25 PM UTC

Retra wrote: post #103

gozo wrote: post #57

Earlier quoted context omitted.

I guess that your opinion. The point is when you change the variables you change the outcome and it won't magically be for the better, especially since there's little precedence. It's pointless to loose karma over this on HN though when I can and regularly do talk about it with actual aerospace engineers.

You could say that about any technology. "The invention of microwave technology will just lead to microwave weapons, Oh No!"

That doesn't make it wrong.

If those "microwave weapons" end up enabling the killing of millions of people (like e.g. nuclear weapons can wipe out the whole earth), then those left will be right to say:

"Hmm, maybe we were better off without microwave technology after all. In fact, just because we can make something, probably doesn't mean that we should make it".