Live data from Hacker News

Data points that Facebook uses to target ads

washingtonpost.com

81–90 of 137 posts

Re: Data points that Facebook uses to target ads

#81

Earlier quoted context omitted.

The worst ads are the ones that advertise the very thing I've recently browsed. Say I browse some shoes (not google them, browse the site selling the shoes), that's when I start seeing tons of ads for that very site. What's the point? I thought ads are supposed to tell me something I don't know?

Nah, the worst ads are the ones that advertise the very thing I've recently bought . Almost every time I buy something on-line from Amazon or some company's store, I then keep seeing ads for that very product . Not before I buy it, but after I've already bought .

Hmm, could there be something in the analytics that could be used to claim, in aggregate as opposed to you personally, that the ads influenced your buying decision, hence demanding a higher price, perhaps for future ads? Coz it doesn't make sense why they'd do that.

Re: Data points that Facebook uses to target ads

#82
post #61

"perfectly targeted ads"? Never had that experience. Even if I summarize all the analytics solutions out there, like ads, Facebook stream, Youtube, Amazon, Netflix. What they usually do is they figure out what I did in recent history or in general and present me more of that. I don't know about other people, but that's absolutely not what I want. That's boring, or maybe even unnecessary. Like, if I just bought a book…

The worst ads are the ones that advertise the very thing I've recently browsed. Say I browse some shoes (not google them, browse the site selling the shoes), that's when I start seeing tons of ads for that very site. What's the point? I thought ads are supposed to tell me something I don't know?

The weirdest ones for me are ads for lumber. Like, I go to the local lumber yard website to check dimensions of what types of lumber is available when I'm planning a project. That lumber is typically costing $1-$2 per yard, so I frequently only buy lumber for ~$20, and they'll show me ads at least ten times. Their ROI has to be very low?

Adding insult to injury, I only buy lumber from their competitor, who, although having a sucky website where I can't look up dimensions, will let me use their compound mitre saw for free, saving me a $600-$1000 purchase.

Re: Data points that Facebook uses to target ads

#83
post #61

"perfectly targeted ads"? Never had that experience. Even if I summarize all the analytics solutions out there, like ads, Facebook stream, Youtube, Amazon, Netflix. What they usually do is they figure out what I did in recent history or in general and present me more of that. I don't know about other people, but that's absolutely not what I want. That's boring, or maybe even unnecessary. Like, if I just bought a book…

Totally agree with this. Even after collecting so much data, Google and Facebook fail at showing me ads which are useful for me. However, even with their inferior targeting FB and Google are making billions. So it might be the case that there is no further incentive to improve in this area and they are focusing on making their core product better to get even more users.

the claimed relevance of the adds is the selling point of google and friends; maybe they do not quite anticipate what a 'power user' might want, however maybe they are more relevant to some people who are not connected day in and day out - sometimes you might do better if you have less data to extrapolate from .

Re: Data points that Facebook uses to target ads

#84

Earlier quoted context omitted.

The worst ads are the ones that advertise the very thing I've recently browsed. Say I browse some shoes (not google them, browse the site selling the shoes), that's when I start seeing tons of ads for that very site. What's the point? I thought ads are supposed to tell me something I don't know?

Nah, the worst ads are the ones that advertise the very thing I've recently bought . Almost every time I buy something on-line from Amazon or some company's store, I then keep seeing ads for that very product . Not before I buy it, but after I've already bought .

I've even seen this for what the companies must know is a once-a-decade purchase, like a big lifetime-warranty espresso grinder. It's not like I need a second one! Offer me some accessories instead, dammit.

Re: Data points that Facebook uses to target ads

#85

Earlier quoted context omitted.

"People who have done things that suggest they're interested in buying this" is a highly valuable segment. Even if half of those people have bought the thing, it's worth advertising to them if it means you get the other half's eyeballs. They don't know that you bought it - they just know that you do things like those who want to buy. Like, I went to a coding bootcamp. Making that decision means browsing the web like…

Oh, but quite often they do know. Amazon knows perfectly well what I bought on their site. Vendor X should know perfectly well what I bought on their own site, using their own store instead of showing me the ad for that product for the next two months.

Even if Amazon knows and organizes the fact that you've bought the thing, that isn't what the ad space is selling. They're selling space that targets a particular kind of person, and whether or not you've bought the thing doesn't matter. Like, the company buying the advertisement slot wants to target a particular sort of person, and this is measured in various ways. So the network selling the advertisement slot wants to group visitors into different buckets based on these measurements, and sell to whoever happens to pay the most for a bucket that view happens to land on.

Basically, it's the end result of a process that confuses a measure with what it's measuring. You want to optimize for "interested in buying X". What you can actually look at is "things that people interested in buying X do more often than people who aren't interested". People buying advertising optimize for that, since it's actually possible to do so. And then since buyers actually care about the measurable proxies, that's what ad-space sellers give.

Re: Data points that Facebook uses to target ads

#86
post #48
post #7

Contrary to popular opinion, people want to see ads. A good ad is very much appreciated. This is why magazines still make money. This is why the SuperBowl is popular amongst non-football fans. Unfortunately, most ads are spammy which is why internet ads have earned this terrible reputation.

I kind of agree with you. What I would love is a system where I could preemptively put in all of my preferences and have some AI go and hunt for things to advertise to me. Like, if I'm in the market for a watch or a car, it knows that, and suggests deals to me. Right now, even during the 99.9% of times I'm absolutely not looking for a car, I have to see the ads, and the ad people lose money on each view. And once I'v…

Yeah, I was thinking about this just today. I'd very happily subscribe to a service to which I volunteer as much data as I want - my age, sex, location, job, interests, whatever - and define categories of products I might be interested in, and such service then shows me non-interactive, text/static-picture ads of relevant products on the web instead of the bullshit we see. An ad-replacer instead of ad-blocker.

Or a lite version, I could get a tailored magazine sent to my mailbox. That would be the single newsletter I'd actually appreciate.

Re: Data points that Facebook uses to target ads

#87
post #71
post #61

"perfectly targeted ads"? Never had that experience. Even if I summarize all the analytics solutions out there, like ads, Facebook stream, Youtube, Amazon, Netflix. What they usually do is they figure out what I did in recent history or in general and present me more of that. I don't know about other people, but that's absolutely not what I want. That's boring, or maybe even unnecessary. Like, if I just bought a book…

Amazon does this all the time with me. "You just buyed a headset, do you want another one?" ... really? How many people buy multiple headsets in rapid succession? Hopefully, their engine will learn with time to differentiate between categories where you often buy many in rapid succession and categories where you only buy every few years, so this gets better. At the moment, it is completely useless.

Well, I bought a headset for home, liked it, then bought an identical headset for the office. Someone else might like it and buy another for a friend. Probably their engine has found out that this is common enough for it to be worth recommending another headset rather than something else.

Re: Data points that Facebook uses to target ads

#88
post #61

"perfectly targeted ads"? Never had that experience. Even if I summarize all the analytics solutions out there, like ads, Facebook stream, Youtube, Amazon, Netflix. What they usually do is they figure out what I did in recent history or in general and present me more of that. I don't know about other people, but that's absolutely not what I want. That's boring, or maybe even unnecessary. Like, if I just bought a book…

> What we need is not a "I know what you did and here is more of that" suggestion engine I really agree on this. The Youtube homepage exemplifies it for me. It recommends only content the same as or very similar to what I've watched before, and only the most popular of that. I've watched a ton of conference talks, programming tutorials, etc. yet those never show up. It's mostly junk with clickbait titles from big cha…

Youtube's home page is especially terrible. It seems to have very broad categories, and there's no really good way of pruning terrible (but popular) stuff out of it. You can say "not interested", then get some almost identical channel instead.

Whereas the "related" sidebar while actually watching a video is great.

Re: Data points that Facebook uses to target ads

#89
post #61

"perfectly targeted ads"? Never had that experience. Even if I summarize all the analytics solutions out there, like ads, Facebook stream, Youtube, Amazon, Netflix. What they usually do is they figure out what I did in recent history or in general and present me more of that. I don't know about other people, but that's absolutely not what I want. That's boring, or maybe even unnecessary. Like, if I just bought a book…

> What we need is not a "I know what you did and here is more of that" suggestion engine I really agree on this. The Youtube homepage exemplifies it for me. It recommends only content the same as or very similar to what I've watched before, and only the most popular of that. I've watched a ton of conference talks, programming tutorials, etc. yet those never show up. It's mostly junk with clickbait titles from big cha…

>And YT does have a page for improving homepage recommendations, which granted I haven't tried yet.

Never hear of it. Can you share a link to it or a name of it?

Re: Data points that Facebook uses to target ads

#90
post #61

"perfectly targeted ads"? Never had that experience. Even if I summarize all the analytics solutions out there, like ads, Facebook stream, Youtube, Amazon, Netflix. What they usually do is they figure out what I did in recent history or in general and present me more of that. I don't know about other people, but that's absolutely not what I want. That's boring, or maybe even unnecessary. Like, if I just bought a book…

Just commenting about Facebook ads, the rest is pretty accurate in my experience -

On FB, that's the remarketing side, and it's a lot more manual. Generally the goals and data will be selected by the advertiser themselves. It works, but it's a bit blunt-weapon-y - let's say it'll catch 10% of people who are absolutely going to convert, and 90% of people who are absolutely uninterested for campaign X. All percentages here are pulled from nowhere, they're just to give you a general idea.

Remarketing is popular because conversions. Also, if you're well-off or otherwise a desirable target, you cost more to reach through Facebook - so you're likely to see more remarketing campaigns, because automatic campaigns will adjust to target less expensive individuals until they've gone through all of them.

Facebook's automatic targeting is a lot more sophisticated, and much more in line with the suggestion engine idea. Let's say it will catch 0.2% of people who are absolutely going to convert, 9.8% of people who might possibly convert, 45% of people who vaguely appreciate the ad being there without converting (including people who might Like/Share/comment on the post), and 45% of people for whom the ad doesn't really connect. It's not usually quite as powerful for conversion, but on average it's pretty accurate at guessing if you'll be interested.

Also, StumbleUpon was a decent suggestion engine once upon a time - if they could actually design a proper extension that did something similar, I'm sure it would be a success.

Post reply on HN