News for hackers. Stuff that matters.
Re:Dutch national broadcaster saw ad revenue rise when it st... by code4tee on Monday July 06, 2020 @10:21AM (#23748173)
I’ve always thought that the “advanced tracking” touted by some in AdTech was:

1. Generally not that advanced (yes I searched for hotels in Florida and so yes I’m probably taking a trip to Florida... don’t need fancy ML to figure that out.)

2. Not that smart. (Hey we actually got back from our Florida vacation two months ago... don’t need to keep seeing hotel offers for Florida!)

It totally makes sense that just “dumb” targeting (show people adverts related to the content on the page they are currently looking at, which requires no tracking) might actually work better than all the black magic products companies promote.

Re:Dutch national broadcaster saw ad revenue rise when it st... by lowwave on Monday July 06, 2020 @10:32AM (#23748278)
It totally makes sense that just “dumb” targeting (show people adverts related to the content on the page they are currently looking at, which requires no tracking) might actually work better than all the black magic products companies promote.

This is exactly kind of "if all you have is a hammer, everything looks like a nail" that turns me off about AI/ML community.

Re:Dutch national broadcaster saw ad revenue rise when it st... by rhizome on Monday July 06, 2020 @07:50PM (#23754106)
Just add 300 DNS queries and wait for the RTB to complete!
Re:Dutch national broadcaster saw ad revenue rise when it st... by harikb on Monday July 06, 2020 @11:33AM (#23748980)
This kind of page-content-based ad targeting did exist way before fine tuned retargeting (say, search or audience/segment retargeting). There are whole bunch of companies that serve this exact market. The reason a more complicated retargeting was created is proof that there is value in it, for advertisers. I am not claiming this is good for people - I am talking only about the business side of things.

> don’t need to keep seeing hotel offers for Florida

This is not for the lack of smartness in code - it is that the conversion funnel is so narrow that "saving that $30 per month, by not showing you the ad, on a $1m campaign" is not worth tracking and excluding. Although I do believe companies should take the extra effort to avoid the annoyance and be a good citizen - particularly once they have identified "the one real person who actually bought something" among 99 other bots/crawlers/untraceable entities. Not all ad companies appropriately value the real person and the annoyance caused to them. Some do.

Re:Dutch national broadcaster saw ad revenue rise when it st... by wastedhours on Monday July 06, 2020 @12:28PM (#23749530)
> The reason a more complicated retargeting was created is proof that there is value in it, for advertisers

I'd argue the value was in it for media buyers. A lot of retargeted inventory is spent targeting people who are going to come back anyway. The agency then spins the numbers and presents it as if they were the only step in the process.

Some even take it a step further and conflate their retargeting with their prospecting numbers - there's a lot more uneducated advertisers out there for agencies to rip-off than educated ones.

Other than your standard ecommerce flows, online advertising is still an area of very murky attribution.

Re:Dutch national broadcaster saw ad revenue rise when it st... by shostack on Tuesday July 07, 2020 @04:05PM (#23763393)
Which is why tools and reporting for incremental lift analysis need to be more widely used. It doesn't help that there are walled data gardens complicating things either.
Re:Dutch national broadcaster saw ad revenue rise when it st... by wastedhours on Wednesday July 08, 2020 @04:07AM (#23767898)
Especially when your dashboards are prepared by the buying agency - always request the raw data behind the scenes. If it takes them a while to find it, there's perhaps a reason...
Re:Dutch national broadcaster saw ad revenue rise when it st... by shostack on Wednesday July 08, 2020 @06:04PM (#23774684)
For sure. Although never attribute to malice what can be attributed to messy spreadsheet exports and databases cobbled together with shoestrings, gum, and vendor swag.
Re:Dutch national broadcaster saw ad revenue rise when it st... by wastedhours on Thursday July 09, 2020 @06:23AM (#23778659)
Oh definitely, but often you're paying them a pretty penny to have it all together, they might not be deliberately BSing you on the data, but they've probably BS'ed you before that they would have the data.
Re:Dutch national broadcaster saw ad revenue rise when it st... by mumblemumble on Monday July 06, 2020 @11:37AM (#23749024)
Ad companies being able to figure out that I just reserved a hotel room, so they can stop advertising hotels to me, might reduce the annoyance level. But I'm also pretty sure that, if I found out that they had done this, I would never patronize that hotel again.
Re:Dutch national broadcaster saw ad revenue rise when it st... by harikb on Monday July 06, 2020 @11:58AM (#23749236)
I am talking about cases where the ad results in a click or view through and finally tracked to a purchase from the same provider. You shouldn’t see the same ad again or another ad to rent a hotel room from same hotel chain. I thought that was the case GP was taking about.

I am excluding the likes of companies like BlueKai (now Oracle) who facilitate trading the cookie to other companies for further targeting or completely different chain of ads based on a search click to another hotel chain

That said, all of this is getting killed by recent Safari and Chrome change of getting rid of all 3rd party cookies

Re:Dutch national broadcaster saw ad revenue rise when it st... by valgaze on Monday July 06, 2020 @11:51AM (#23749159)
@harikb, what about something like this: https://i.imgur.com/LYU6NBb.png

Surrounded by content on advice "blue light"-- gets targeted for Blue-tooth lightbulb

Re:Dutch national broadcaster saw ad revenue rise when it st... by Traubenfuchs on Monday July 06, 2020 @12:41PM (#23749672)
Amazon showing me a selection of typical buy-once products of the kind I just bought is a depressing sight. They have infinite money and his is the best they can come up with?
Re:Dutch national broadcaster saw ad revenue rise when it st... by grandmczeb on Monday July 06, 2020 @12:48PM (#23749738)
I don’t understand why this is such an obviously bad strategy - lots of people buy an item, cancel or return it, and then buy a similar item.
Re:Dutch national broadcaster saw ad revenue rise when it st... by cgriswald on Monday July 06, 2020 @01:04PM (#23749910)
Let’s say you buy a washing machine on Amazon. You’re not likely to need more than one. Amazon knows whether you canceled or returned your order.

So the only way it makes any sense is if Amazon is advertising on the chance that you’re going to cancel or return it in the future. Amazon certainly has those numbers. But do you think the chance you’re going to cancel or return it is so high Amazon should advertise the same product to you instead of something else you might buy?

Re:Dutch national broadcaster saw ad revenue rise when it st... by grandmczeb on Monday July 06, 2020 @01:32PM (#23750321)
> But do you think the chance you’re going to cancel or return it is so high Amazon should advertise the same product to you instead of something else you might buy?

It’s not either or - Amazon showed me grill accessories along with other models when I bought a grill. Regardless, I don’t really have a hard time believing that even a small chance of buying and canceling could make the expected value of showing that item higher than other items, even if that’s counter intuitive to you. I have a harder time believing that what’s shown is not well optimized and “obviously” wrong.

Re:Dutch national broadcaster saw ad revenue rise when it st... by danaris on Monday July 06, 2020 @02:40PM (#23751216)
Any individual spot Amazon is advertising you something they could easily know you will not buy is a spot they could have instead used for an ad for something they do not have this knowledge about.

So yes, it is either/or.

Re:Dutch national broadcaster saw ad revenue rise when it st... by grandmczeb on Monday July 06, 2020 @03:55PM (#23752014)
The point is that if you have multiple ad slots you can cover a wider range of buyers. For example, the chances that someone would click on a 5th grill accessory ad is probably pretty low if they didn’t click on any of the previous 4. But if they’re thinking of returning the grill they already purchased, they might be interested in a different model. Averaged over all buyers, the best strategy might well be to show a mix of accessories and other models.

Or, depending on the numbers, all grills or all accessories could also make sense. My broader point is that it’s perfectly possible the best strategy doesn’t line up with people’s intuition, which is why we use ML for recommendations in the first place.

Re:Dutch national broadcaster saw ad revenue rise when it st... by danaris on Monday July 06, 2020 @08:05PM (#23754218)
I mean, sure, I don't have any scientific studies to back this up.

But I really don't feel like this is a case where "intuition might be completely wrong" applies—buying a $400 grill, and then just going back and buying 3 more $3-500 grills, just isn't something that people do on a regular basis. Same for vacuum cleaners, refrigerators, etc. And yet Amazon does regularly advertise these things to us after such large purchases.

The only logical conclusion is that it's cheaper for them (in the short term, at least) not to update their ad/recommendation algorithm to take into account the nature of large one-time purchases.

Re:Dutch national broadcaster saw ad revenue rise when it st... by grandmczeb on Tuesday July 07, 2020 @08:10PM (#23765713)
The question isn’t “do they do it on a regular basis” it’s “do people do it often enough that the expected value of showing that product is higher than something else.”

> The only logical conclusion is that it's cheaper for them (in the short term, at least) not to update their ad/recommendation algorithm to take into account the nature of large one-time purchases.

Is that really the only logical conclusion? I’d think that “I don’t know people’s buying behavior on Amazon well enough to understand why this happens” should at least be a possible explanation.

Re:Dutch national broadcaster saw ad revenue rise when it st... by cgriswald on Monday July 06, 2020 @05:38PM (#23753089)
I'm trying to imagine a scenario where someone returns a single-purchase item and then purchases the same exact item as the result of seeing an ad for it. They're already aware of the item and if they want another one after return they're probably just going to exchange it because something is wrong with it. It's not even a sale and I'm skeptical that seeing the ad has a strong influence on whether a customer asks for a refund or a replacement.

ML is not some holy grail that always produces optimum algorithms. It depends heavily on inputs and on the assumptions, both explicit and implicit in the training data. My guess is Amazon does not have a "single-purchase item" model and it's not something the AI was trained to identify and so they aren't specifically excluded from groups of "similar-or-same" items.

I think it is far more likely that Amazon is optimizing the cost of development against the marginal benefit of not showing ads for single-purchase items than that this is somehow driving more sales.

Re:Dutch national broadcaster saw ad revenue rise when it st... by grandmczeb on Tuesday July 07, 2020 @08:01PM (#23765657)
I’ve personally done the following:

  1. Buy a large ticket item (e.g. furniture).
  2. See an ad on Amazon for a different/nicer/better/faster shipped version.
  3. Buy it.
  4. Cancel the old order.
Is that a common pattern? No idea. But it’s not crazy to me that it happens often enough that it’s worth it to Amazon to show similar items to people who already bought them.
Re:Dutch national broadcaster saw ad revenue rise when it st... by Skunkleton on Monday July 06, 2020 @05:08PM (#23752773)
Or even better: show a grill accessory, but also show some other stuff perhaps unrelated to my new found love of buying and returning grills?
Re:Dutch national broadcaster saw ad revenue rise when it st... by marblessubtext on Monday July 06, 2020 @02:30PM (#23751089)
I bought a quality vacuum from Amazon and then continued to be aggressively advertised vacuums for months (on Amazon still) while shopping for things that honestly I could have used some advertising for to let me know what even existed in the space. Seems like not a great strategy for any involved party.

They never advertised bags for said vacuum...

Re:Dutch national broadcaster saw ad revenue rise when it st... by frogpelt on Monday July 06, 2020 @01:26PM (#23750230)
Would you say the majority of people shop like that?

It makes sense to me that if I buy post hole diggers you should show me ads for work gloves or mailboxes or fences. Not more post hole diggers.

Weird example, I know.

Re:Dutch national broadcaster saw ad revenue rise when it st... by jacobr1 on Monday July 06, 2020 @03:27PM (#23751737)
That would probably be a better strategy. But if you are looking at a naive, single variable, a prior purchase is the greatest single predictor of future purchases. If you know nothing else about a person, the fact you purchased one post-hole digger makes it more likely you'll purchase another, as opposed to some random other internet denizen for which we know nothing. Having worked in email marketing analysis before, this effect was surprisingly robust across categories. It implies you are the kind of person that could use such a product - most people aren't.
Re:Dutch national broadcaster saw ad revenue rise when it st... by grandmczeb on Monday July 06, 2020 @01:41PM (#23750455)
I mean, I’ve definitely done it and I’ve seen other people do it. I have no idea how typical it is.
Re:Dutch national broadcaster saw ad revenue rise when it st... by oh_sigh on Monday July 06, 2020 @01:18PM (#23750101)
The worse is they keep showing me, say a stereo receiver even after they know I bought one, and it makes me feel like maybe I bought the wrong product when I see other products with better features/price advertised. These ads literally give me buyers remorse for the product that they themselves sold me.
Re:Dutch national broadcaster saw ad revenue rise when it st... by afiori on Monday July 06, 2020 @10:30AM (#23748263)
The best argument for the usefulness/profitability of deep tracking is simple statistical inference.

As in if there are 10000 people that with browsing patterns similar to yours and non-trivial fraction of them one day show interest in a discount/service/trip then you are also likely to be interested.

It is the official explanation of how facebook/google can magically listen to conversation IRL that they were not supposed to hear: we simply are not as unique/unpredictable as we would like to think.

Re:Dutch national broadcaster saw ad revenue rise when it st... by baybal2 on Monday July 06, 2020 @10:56AM (#23748590)
I think it's one of the best argument against.

It was at least 5 years ago when I started to notice businesses "A/B testing themselves to death"

— "It makes no common sense, but increases sales,"

and then it does until it doesn't, and you are left with no idea what your client actually wants, and no idea how to get out of that "A/B tested to absurdity" situation

Re:Dutch national broadcaster saw ad revenue rise when it st... by cortesoft on Monday July 06, 2020 @12:00PM (#23749251)
Yeah, I often wonder this... a/b tests only ever measure short term differences. What if the change you are making increased sales for a few months, but pisses off users in the long run?

An A/B test isn't going to show that.

Re:Dutch national broadcaster saw ad revenue rise when it st... by shostack on Tuesday July 07, 2020 @07:31PM (#23765445)
A regression model might if you have sufficient data.
Re:Dutch national broadcaster saw ad revenue rise when it st... by cortesoft on Tuesday July 07, 2020 @10:09PM (#23766268)
You would have to run the a/b test for quite a while, and I don't think many companies do.
Re:Dutch national broadcaster saw ad revenue rise when it st... by mumblemumble on Monday July 06, 2020 @11:51AM (#23749160)
This also seems to be an inherent danger with thinking that data science == machine learning == AI.

I don't want to fundamentally challenge the ŷ-centric approach approach of machine learning. But I do want to say that there's a lot of danger in taking it too far, and thinking that β̂ doesn't matter as long as ŷ is good enough. β̂ has a habit of not mattering until, very suddenly, it really matters quite a lot. And it's even more dangerous to think that an algorithm for generating formulas is somehow doing something intelligent just because you like what you're seeing in the ŷ department.

Re:Dutch national broadcaster saw ad revenue rise when it st... by afiori on Monday July 06, 2020 @12:09PM (#23749334)
As someone that is not "in" with those fields what do ŷ and β̂ stand for here?
Re:Dutch national broadcaster saw ad revenue rise when it st... by mumblemumble on Monday July 06, 2020 @01:23PM (#23750183)
ŷ is the prediction you get out of the formula, β̂ is the coefficient(s) on the variables in the formula.

So, take the simplest possible model: You think that your dependent variable is linear in the dependent variable, and that the line passes through (0,0). That gives us a simplified version of the classic equation for a line, where we say that y is equal to x times some constant that indicates the slope of the line:

  y = mx
In machine learning (or statistical inference), the variable conventions are slightly different, so it would look more like:

  y = βx
And then, since this is machine learning (or statistical inference, or whatever), we don't know the true equation, we're just working with an estimate we've inferred. We stick hats over the estimated values to make it clear that they're estimates:

  ŷ = β̂x

What I was describing, then, is which side of the formula you're focusing on. Traditionally, statistics has been all about scientific inference, and trying to figure out and characterize causal relationships. So you want to know "If I change x in some way, how will that affect the outcome, all else being equal?" β̂ is your tool for teasing those things apart. Machine learning is typically cast as being more about making predictions from observational data. ŷ is just the formal notation for describing those predictions.

It's pretty common in machine learning to say, "We are working with observational data, we're not worried about causality, so we'll only pay attention to the predictions." There can be value to that approach, but only to the extent that you really don't care about causality. Which. . . surprisingly often, β̂ matters, even when the people who built the model were trying to ignore it. It becomes especially important when you're using the model to make decisions. For example, when a company is using a machine learning model to try and identify candidates for a position they're trying to fill, they should absolutely be paying attention to what the model thinks about people who liked Lane Bryant on Facebook, and how that compares to people who liked Jos A Bank, and whether they really want things like that to be playing into how they make hiring decisions.

Re:Dutch national broadcaster saw ad revenue rise when it st... by code4tee on Monday July 06, 2020 @11:11AM (#23748744)
I see that but I think the broader point (and the point of the article) is that too often “success” in finding some less obvious link actually comes at the cost of overlooking the simple stuff. In the end there’s an argument that while neat the fancy ad targeting actually fails to meet the core business objective of maximizing ad performance.
Re:Dutch national broadcaster saw ad revenue rise when it st... by afiori on Monday July 06, 2020 @12:11PM (#23749350)
My personal opinion is that for most companies having an organically grown audience is more valuable than a lot of faceless traffic.

Obviously they can help each other if you do things right and are lucky.

Re:Dutch national broadcaster saw ad revenue rise when it st... by Swizec on Monday July 06, 2020 @10:40AM (#23748375)
My girlfriend is currently experiencing the dark side of smart tracking. Female who hit 30 and is in a stable relationship.

The ML just can’t figure her out. One day every ad is for dating sites, the next day it’s all babies and pregnancy, then engagement rings, wedding locations, more dating sites, a few days of houses to buy ... it’s like the algorithms can’t decide what life stage she’s in.

Re:Dutch national broadcaster saw ad revenue rise when it st... by harry8 on Monday July 06, 2020 @11:13AM (#23748770)
That's what you want. You don't want these assholes knowing your business. The classic example is when they start advertising a whole lot of pregancy stuff to someone who doesn't want to be pregant and for whom that might be a serious personal problem - advertising seen by others who she does not want to see that.

Dumb algos means freedom and privacy. Smart algos that actually work stuff out are dangerous and vile. Nobody gave them permission for that. Nobody complains about seeing ads not specifically tailored for them.

Re:Dutch national broadcaster saw ad revenue rise when it st... by lmkg on Monday July 06, 2020 @11:32AM (#23748977)
Confounding all of the targeting is one simple fact: Ad selection is based on what the advertiser is willing to spend, not on what the consumer wants to see. If a jewelry shop has extra ad budget and their target market is saturated, then the ad network starts serving ads to people who are married. Either that or they tell the jeweler to stop giving them money.
Re:Dutch national broadcaster saw ad revenue rise when it st... by Traubenfuchs on Monday July 06, 2020 @12:43PM (#23749692)
Watch out... Maybe the algorithm knows EXACTLY what she is thinking during any given surfing session?
Re:Dutch national broadcaster saw ad revenue rise when it st... by ceejayoz on Monday July 06, 2020 @10:55AM (#23748567)
> it’s like the algorithms can’t decide what life stage she’s in

Or, it's working on figuring that out.

Re:Dutch national broadcaster saw ad revenue rise when it st... by phkahler on Monday July 06, 2020 @11:36AM (#23749012)
Just like she is right? Are you going to marry this guy or move on? Are kids in your future? How about a house, are you two moving in together? Probably all the same crap nosey family and friends want to know. Are advertisers an omnipresent nosey entity? Probably :-)
Re:Dutch national broadcaster saw ad revenue rise when it st... by synthc on Monday July 06, 2020 @11:13AM (#23748769)
When my first kid was born, all the ads I got where either baby clothes, daipers etc., or Second Love. A clear separation between A and B
Re:Dutch national broadcaster saw ad revenue rise when it st... by mcv on Tuesday July 07, 2020 @08:27AM (#23758453)
Makes me wonder if there are a lot of people who start cheating once they get a child. What exactly is that algorithm trained on?