So there are essentially remnants of our browsing history linked to our devices shared among numerous ad companies.
They then serve relevant ads for us all over the web depending on where they are being paid to display relevant ads.
Twitter is at it. We've experienced the same behavior from Google & God knows Facebook is at it too.
I've even had conversations where the only connection we had to the web was our locally running Alexa only to see ads relating to our specific conversation 10 minutes later on the web.
Can anybody think of a technological approach to flagging this behavior?
The difference between targeted ads and untargeted is like the difference between alpha and beta, or index funds and hedge funds. It’s not that a hedge fund doesn’t have better risk adjusted returns than an index (they usually do), it’s that the fees are so high it’s usually worse net of fees.
There was a study posted on hacker news that targeted ads have something like a 5% higher conversion rate than untargeted ones. The problem is that you end up paying 30% more compared to the commodity service. So unless the fees go down or the efficacy goes up, it’s not worth it.
I'm confused. If this technology is so pervasive and supposedly efficient, why do I always get such terrible ads, completely irrelevant to my interests? Usually the only targeted ads I get are ads for products I was searching for to buy, but usually they show up after I bought the product, so, no, I am not going to buy a new fridge within few years anyway.
This happens for the same reason a close friend or a family member gives you a terrible Christmas gift. Ostensibly they know all about you, your values, and your preferences. Despite this, it's actually hard to predict what someone really wants or desires.
Not only does this problem remain true despite the high amount of information that advertisers have, but it's compounded by the business model. They don't want you to buy just anything you'd like, they want to sell you some particular product or service which you may never like, no matter the information that's been gathered on you.
There's a lot of wasted inventory yeah, but conversion rates are still much higher than they would've been without targeting. It looks terrible from the individual's point of view but from on high it seems to work well enough.
By conversion rates, do you mean clicks or sales? The CTR industry (and by that I mean people who "optimize" landing pages and ads) count clicks- but as the person paying for it I count sales and don't care about clicks.
Conversion means whatever the end company wants it to mean. If it's mass retail that probably means sales (like in dollar value or whatever). If it's something else it might mean account creation. If it's a very high touch sales thing it might mean initiating contact with a human sales person. And yes, it might mean clicking through to a specific part of the website (account creation is a subset of this).
Part of a successful as campaign has to be defining your metrics correctly. The closer to your actual end business goal, the more realistic your performance tracking is.
- Confirmation bias. You remember the irrelevant ads, because they made you do a double take, they stood out as unusual.
- The terrible ads are actually effective at getting clicks from people with similar interests, but less of an aversion to ads, less ad-blindness due to depression, more willing to buy online, ...
- Especially with retargeting (what you describe with the fridge), the CTR is higher than with non-retargeted ads. But it is also one of the most visible forms of advertisement (everyone has a story similar to you, where they get followed by ads after looking at cars or fridges).
The online ad industry fuels the likes of Google and Facebook. Tremendous engineering and research effort goes into making these systems more efficient. Personalization and tracking of preferences is at the state of the art for these companies. Though not the potential best, it is close to the best possible we can do with ML right now.
I'm struggling to find my sources, but I read years ago that most advertisements for cars are not to encourage more people to buy a new car, but are actually aimed at people who already own the car. Seeing their purchase being advertised on TV increases it's relative value to them.
Ultimately they feel affirmed in their purchase, overvalue their "asset", and increase in brand loyalty.
Because the way that adtech takes credit for sales has very little to do with whether it actually influences you. So they claim to be effective at the same time as doing the stupid stuff you point out.
It’s crazy that ~10 years ago I tried a pirated copy of Dragon NaturallySpeaking on a 700mHz single-core machine with 384MB of RAM and it was capable of recognising my speech locally but now they want us to believe that local speech recognition is impossible despite having 10x the processing power.
I thought with Alexa we were still in the "in some dystopian future they're going to use that data for ads" phase. If we've moved to that already, I'd like to read about it.
It's not a technological problem. Only laws supported by painful enforcement could stop the flood of shit that adtech unleashes on the world. If you promise people to torture them with ads and sell all their data they'll come in droves because they get something for “free”. Some people even pay money for it.
Why would you have a surveiliance node in your home if you are concerned about privacy?
As for the other thing, I am running this in firefox with containers active. Google gets their own, Facebook gets theirs, twitter gets theirs and mostly everything else gets a temporary container. There will be no connection between them to use my data on, and in a few minutes my records of their cookies will disappear.
This is conspiracy theory nonsense. If they could do this they'd be selling it (Amazon doesn't give away stuff like this for free) and people would be lining up to buy it.
Amazon's code is not secret internally. If Alexa was targeting ads on your speech, someone from Amazon present or past would have confirmed it by now. Your conversation was probably influenced by the web ads campaign, or it was a coincidence, there are a lot of people seeing a lot of ads every 10 minutes.
Amazon's code is not secret internally. If Alexa was targeting ads on your speech, someone from Amazon present or past would have confirmed it by now.
Or not? How many people have access to Alexa source code, let alone that part of Alexa source code, and are also willing to break their NDAs and get fired and/or sued?
I like your point as I can relate to that opinion. However my outlook has changed in the last year or so.
There has been several occasions where my conversation somehow turned into relevant web ads. I don't think I'm making a mistake here.
A friend of mine told me of his good friend who left Amazon recently as his job was to listen to what users say to Alexa and ensure Alexa responded correctly. FYI that's both when users say Alexa first and also when they do not. More on that here [1].
I would love some more clarity into what's going on here.
1) I would assume your entire social network is well documented. A lot of conversations include one person looking something up before or after the conversation. Most conversations don’t come up randomly. That can probably signal that you might be interested in such a topic too.
For 2) it’s worth noting that when you say you didn’t say Alexa, it means Alexa at least thought you did. It’s not trying to constantly record, but may still be a problem to you.
I've been keeping my eye out for good, solid proof that these assistants are doing ad-based targeting based on things being overheard, but I haven't found anything despite looking, and the arguments they aren't are reasonably compelling. (The people looking are quite motivated to find it. Big feather in their hat if they do and can even convincingly cast shade on Amazon, let alone prove it.)
My conclusion is that the targeted ads that people report are either A: coincidence (millions of users, only really tens of thousands of types of "things" to advertise with, sooner or later life coincidences would be expected to happen to large numbers of people) and/or priming (you notice the ad for camping gear since you were just talking about it, but don't notice the several hundred other impressions for camping gear that you just edited out entirely) or B: they didn't get it from your speech; you gave out other signals without realizing it, like a Google search that may not have been "wanna buy camping gear" but unknownst to you is known to be correlated to it, or as you say, your friend ran a search and now you're getting ads, or you're talking about camping gear in the first place because you are in a coffee shop near Camping World Outlet, and in general, the advertising machine knows that you're interested in camping gear right now because that's how well they have you under surveillance... they don't need your actual speech to know that's what you're talking about.
The first is mathematically inevitable, but poorly explains the cases where it recurs to one person. That seems to be the second. In a sort of way, I say that's even creepier than "they're listening to your literal speech and targeting you based on that". They've got you so pegged from everything else they're doing they don't even need that signal. That's creepy.
People have dozens of conversations every day that don’t match with the hundreds of ads they see every day. But then once or twice coincidentally a conversation preceded a relevant ad and the user jumps to the conclusion that their conversation was being listened to without noticing that 99.9% of the time the ads have nothing to do with any topic of conversation. Its occam’s razor or confirmation bias or some other phenomena HN likes to point out.
Alexa does not send data unless you say “Alexa”. This happens at a hardware level. This is confirmed by hardware schematics and by security researchers who have sifted through every piece of data that leaves the device.
How reliable is the keyword detection? Last week we had this story about Apple where Siri heard the word "Siri" out of random noise like creaking furniture.
the device can't magically detect the world Alexa - there is some lossy process of translating incoming noise into a vector that has some probability of being the hotword, and above this probability the device enables.
a more accurate thing to say would be:
Alexa does not send data unless, in the course of listening for the hotword Alexa, it interprets some noise or utterance to be close enough to Alexa to start sending data
Whether or not they are actually listening to you is a detail. The creepy thing is that they obviously know you as good as if they were listening. How that is achieved isn't really the point.
They then serve relevant ads for us all over the web depending on where they are being paid to display relevant ads.
Twitter is at it. We've experienced the same behavior from Google & God knows Facebook is at it too.
I've even had conversations where the only connection we had to the web was our locally running Alexa only to see ads relating to our specific conversation 10 minutes later on the web.
Can anybody think of a technological approach to flagging this behavior?