alt Hacker NewsI wish tech companies knew more about me
Comments
Google's real advertising proposition is they will sell you the audience of an expensive website on a cheaper website. Want to advertise on the New York Times? Why pay their high rates? Google tracks who visits the NYT and offers the advertiser the same audience when they visit Bob's hobby gardening website, which is considerably cheaper to advertise on. That's really why they want to track where you go. Understanding what you want is just a secondary aim to that, if they are interested in it at all, and I agree with the article that they are very poor indeed at it.
Targeted advertising will never feel good, and it's not due to lack of information. Consider the following:
- People who know you very well (friends, family, spouse) often buy you bad gifts you need to pretend to be happy to receive.
- Ultimately, advertisers aren't _merely_ trying to find things you want. They're trying to get you to buy _their_ things. So someone selling (let's say) hamburgers might be savvy enough to avoid vegetarians, but after that how would they possibly narrow it down? Not everyone wants to try out a new burger place, or even cares about advertising when it comes to restaurant suggestions.
ie, they're trying to coerce you into wanting their products, and no amount of targeting can really fix that problem. Also, targeting is difficult.
On occasion I get a glimpse at someone else's experience on the web and I'm so thankful that I haven't seen many ads at all in years.
I just disabled my adblocker on the BBC website out of curiosity and oh am I happy I don't live like this.
It is a hard truth to swallow for many on HN (myself included) but the reality is a lot of normies don't mind ads at all. I recently told my wife how to install an ad blocker and her response was "but sometimes I see cool stuff in ads". I'm sure the percentage of "cool stuff" she sees is higher because of what they know about her.
Amazon (or whoever) showing you the same stuff you just bought is kind of a separate issue, because it obviously isn't useful to anyone (including Amazon) and surely is so easy to avoid I find it bizarre that it still happens in this day and age.
As an aside, there is exactly one place on earth where I will actively look around for ads, and that is London Underground stations. Because for whatever reason they advertise a lot of cultural events (concerts, festivals, plays, etc).
Yup, it's weird to me that companies as big as Amazon haven't figured this out.
Guys, if I buy a photo camera, I might want lenses, or a cleaning kit, or maybe a camera bag. But I'm almost certainly not interested in starting a camera collection, nor am I interested in incompatible brands to what I just bought.
It shouldn't be so hard to run some stats and figure out that there's things one doesn't buy again often, like washing machines, cameras, laptops and other big purchases.
Also, search still sucks. Searching for things like sizes where fuzzy ranges are involved is pain. Like I need to find a recessed light that fits into a 100mm hole, it's a pain to search for still. At this point I think somebody at Amazon could have come up with a way to categorize products and extract parameters automatically. And the pain only increases exponentially if you need 3-4 things in a product to match.
ChatGPT can do it, no reason why Amazon couldn't just have that on their page.
It actually shocks me how often technology in general does the exact wrong thing to what I want. A couple off the top of my head.
* I purchase something and all the recommended products become the thing I just purchased for the next month.
* I do an operation in 3D modeling software that can be applied in 2 directions, it always chooses to apply it in the direction that makes no sense.
* I get recommendations in my social media feeds for what to do in a certain city for 2 months after I already visited and returned home.
* I click on a link to a song, listen for 10 seconds and turn it off because its a genre I don't like, too bad, that is what the algorithm thinks I'm exclusively into now (related, right now I'm struggling with Tidal recommending Swedish language songs because a band I recently discovered is Swedish despite singing in English).
Maybe I don't notice when the computer gets it right most of the time, but it really feels like it gets it wrong more than it gets it right. It also feels like this continually gets worse, when software should be getting smarter at predicting what a user wants. Is it unreasonable to expect that my computer can see that I did the same operation on something twice in a row and that maybe it should prompt me to do it on the rest of the data in the same location?
The flipside of this was a shoe-shine boy moment for me about ads and advertising, recently and personally.
A small storefront business I liked closed up over the summer. Literally nothing to do with tech other than they had a website, flat HTML that hadn't been touched in years, to solicit business. They communicated with me and their other customers by SMS, almost exclusively.
I talked to one of the core employees and asked why they were closing. Business was drying up, the usual story of business headwinds but among the reasons business was drying up was they couldn't make money by running ads, the cost of running the ads these days meant they would barely break even on the average customer they acquired at the historical customer acquisition rates.
I don't buy ads myself but the sense they shared was the like the locksmith and plumber experience, except as seen from the other side. Maybe it's different in other parts of the world but the experience is 1000s of ads for indistinguishable fly-by-night-looking operations all of which seemingly lead to the same call centers. Dispatching nearly-faceless vans from who knows where and a different junior-ish guy or gal shows up each time, looks dazed, does an average job and hustles off to the next one.
because you're not the client.
advertisers are.
you use adblocker. you're persona non grata to begin with. your purchases are not mapped back, that's why you keep seeing the stuff you just bought. for them, you are top of the funnel still. AdBlockers work! (20yrs in adtech)
for music, you're also irrelevant. you're old. music industry doesn't care about 25+ finding new music. they spend all their effort on keeping teenagers in a comfortable bubble to mine them for music purchases and shows. people under 25 spend pretty much all the money record labels make.
anyway, the system works and they do know everything about you. but your goals and desires are completely irrelevant to everyone involved money changing hands.
Reminds me a bit of Paul Miller's UI Dyson Sphere concept: https://www.theverge.com/2011/08/03/535880/natural-user-inte...
Basically: capture everything I'm putting out and use it to help me.
This! I don't mind tech companies knowing things about me. If I don't want them to know, I browse in Safari's anonymous mode.
But I hate it when they suggest things that I don't want.. show me stuff I have already bought.... show me things I already decided against... show me stuff that is not in my price range. That's annoying. I don't mind a good, paid recommendation. I just don't like all the bad recommendations.
Yeah, the fact that too many websites just ignore language settings/preferences (both system, and browser it seems) is really annoying. Location-based language picking doesn't even make sense in some locations. Like ebay not even giving an option to switch from German to anything else, if you are in Switzerland. They seem to be unaware that Switzerland is a multilingual country?
They also can’t really tell apart which product are recurring buys.
If I buy toilet paper and a toilet seat I’m not going to be buying another seat next month. I don’t use them up like that.
Always surprises me that Amazon hasn’t managed to incorporate that into their algo
All those ads where paid by some companies, who were probably promised by some google ad salesperson the most focussed and personalized and relevant targeting. Will they get a discount ?
Can the advertiser (as in "the company paying for the ad") get any of those signals ("we're paying for ads, but they're being shown to randos who are never going to by our products") ?
Or are the online ads so cheap that they don't care ?
What the author wants isn't for tech companies to know per more, but use what they know in per interest. Ads are shown for products you just bought because the algorithm is optimized to maximize conversion of clicks to purchases, not to please the buyer which is a subtle but important difference, and this is what this optimization looks like.
> I get shown ads for products I just purchased. All this spying, machine learning, big data result in "he bought a mattress, show him more mattresses".
You know, I had a really long argument about a decade ago with an Amazon senior engineer on the ads side of things about this. I was complaining about repeatedly being shown ads for the same blender I had just purchased (not accessories, not competing products, the exact same blender). And I hit a brick wall of "everybody complains about this, and we've been trying to change this, but every change to this makes the metrics worse".
The other bizarre case is how there's no accounting for social class: it ought to be relatively easy to infer, yet they still serve ads that you'd never ever consider, even though there are ones you might (at a push: I'm generally keen to avoid ads always)
> I have no doubts that a lot of information is being tracked, stored, and sold. Since that's already done, could you guys at least use some of it?..
You can be certain that they are using that data. The truth is that the data being collected about you is not intended to benefit you, it's being gathered and stored and sold to be used against you in any way that the person with that data feels might be an advantage for them.
It's not really about showing you relevant ads, or helpful recommendations, or useful product offerings. It hasn't been about any of those things for a very long time now. Your data will still be used that way, but while companies will tell you that is mainly why they collect your data that is a lie. Those are secondary uses for your data and making those things work well for your benefit is a cost to the companies and they are only willing to invest so much time/money/effort into that.
Your data is being used mostly for their benefit at your expense, and where it isn't being used today, it's being stored for a time in the future when it can be used more effectively against you. AI promises to make it easier than ever to trawl through all that data and find ways to leverage it to better manipulate you, or take more of your money from you, or further limit any possible risk you might expose a company to.
Companies don't want you to be aware of when and how that data is being used, so it's natural for you to feel like it's all being ignored, but you can sure that a lot of work is going into finding new and creative ways to put all that data to use.
I agree with the author that it is super annoying when you are assaulted with ads for the things that you have bought recently.
If you have bought them recently, they are the most unlikely things to buy in the near future.
I was actually surprised a few days ago when I have seen for the first time in many years a personalized ad that was useful. When I visited an on-line shop, I was greeted with a message that a product at which I have looked during a previous visit, but which I have not bought, has a 20% price reduction, and at the new price I was indeed interested to buy it.
This kind of notification, about a price reduction, makes sense, but ads for products to which I have looked previously and nothing has changed about them do not achieve any useful purpose. If I bought them I have no need to buy them again, if I did not buy them, this means that I decided that I do not need or want them, so I will not buy them regardless if I see ads for them.
I agree with parts of this in the sense that I find autoplay on Spotify and many machine recommendations quite bad.
However:
a) I don't want better-targeted ads or more convenient infinite recommendation algorithms because they create friction that preserves some intentionality in my media consumption.
b) there's some survivorship bias here. We notice the bad things because they stand out, but not all of the machine-recommended things that we unquestioningly accept.
Goes to show the middlemen (the ad engines, ad auction sites) don't really have either the user or the advertiser's best interests in their mind. They just want to fill up the slots, and some ad agency is churning out ads and all they care about is some CTR/CPC metrics to get their fees, and the end manufacturer adds that markup to the product and charges them to the actual legitimate customers of those products.
Not even tracking, but when I land on a page from a Google search, some sites try to be helpful by steering me to a local site, but the local link does not exist so I end up on the homepage, not the specific help I was looking for.
Adobe and Lenovo are really good at this.
the horrible performance of targeted ads in the op for both consumers and advertisers is why many advertisers are now going for connected tv, or native ads in market places (e.g you're shopping for detergent - then you see a sponsored listing for another detergent brand)
coz targeted ads as in op - only one party has benefited so far - people running the ad exchanges.
Recently I had my AIs do a massive deep-dive into all traffic going into and out of my phone. The amount of analytics and data being collected is absurd. Even the Android launcher app I use (Nova) is essentially controlled by an analytics company. Everything you do, even on your home screen, is sent to their servers.
A very odd stance. Personally if I'm going to be tracked and advertised to I want it to be things I have absolutely no interest in, even better if it's in a language I don't speak.
I don't think the kind of problem that advertisement tries to solve, in our minds as users, can be solved. You can't possibly predict what I will want in the near future unless you can predict my future evolution in tastes and needs, which is a very hard problem in my opinion. Advertisement is needed because the next time you will by chance want a certain kind of product, you will maybe remember a brand over another.
The article assumes adverts are there to show them relevant products that they actually might buy, but adverts are there for the platform to make money. The result of this is that they cast the net far and wide and say you are 'a targeted user' so they can push the advert to you and make more money.
Loved that little rant. You had me at "I skip every single electronic track with vocals".
I feel the same way about Youtube's recommendations. They always end up showing me the most bland milquetoast videos you can think of, videos seemingly trying to bait me into watching them, or videos by people who, to be fair, are somewhat similar to people I already watch consistently, but who just don't jive with me. Clicking the 'not interested' button is a non-signal, just how subscribing apparently just isn't a signal they use as well.
I could probably get something pretty decent going if Youtube would give me access to some buttons that actually give their algorithm signals directly. Subscribing and watching every video should be a pretty strong signal. Subscribing and only watching videos with a certain subtitle should be a signal of what videos on that channel I like. There should be a 'channels I don't really care for' bucket, that it might still show me videos from if it's doing exceptionally well. But giving users these tools isn't in the interest of Youtube, so they never will.
I do much better after switching to FreeTube as a front-end for Youtube, since now I'll only be shown videos from people I subscribe to, and I can be on my own to find videos I actually want to watch. If only Youtube search wasn't hot garbage, I might actually see success more often than I do now.
I have enough ad and tracking blocking that the ads I get are totally misdirected. I get ads for drugs for diseases I've never heard of. I get ads for equipment for processing and bottling olive oil in bulk, despite being nowhere near that industry. I get ads from China for very large hydraulic presses, despite not having any great need for metal-forming equipment. And World Dairy Expo wants me to attend, although I don't have any cows.
So needed a new toilet seat. My middle aged hairy booty broke the one I had finally. Now Amazon thinks i want toilet seats on subscription. Toilet seat of the months etc. Buy it again... toilet seat reviews even... I wonder if it's not amounting to how much tech companies 'know' but more how much have they automated 'caring'
I think it'd be a great opportunity for someone to make an ad network where the user can type in things about themselves in an ad space. Randomly targeted questions and then we get ads perhaps around that. Ads that start off about current events in a fun way and give users an opportunity to feel like they have control in what they see. I know users would abuse it but I think that'd add value to the interaction. If I didn't want toilet seats but I write that I do it'd be valuable to an ad network to see how users interact with their preferences. I'm sure we'll say google already sort of does this but I disagree. I think google infers based on our traffic. Some sites I visit I don't necessary want to but am forced to. I'd love to see ads or have the option to even pay for things I want to see like pictures of my kids instead of an ad.
And Google cannot correctly classify the spam I'm receiving. For example:
Subject: 6MONTHS OF'COVERA.GE'START.ING.AT $2
From: RR_BANKAI.NET.bankai.NET.MCK.NET.MCK.NET.MCK.NET.RR_BANKAIdb@ad288.com
The language thing shits me to tears.
I lived in Germany for years but don't speak German. I raised a support ticket with Spotify to ask that they stop sending me emails in German, as they full-well know I speak English (the Spotify UI, thankfully, is in English). They declined. I sent their emails to spam.
Amazon is the same. I ended up just not using them, because it was so painful getting their site to talk to me in a language I understand.
And, as TFA points out, we have a mechanism for this: the browser passes a preferred locale to the server. It's simply a matter of the server respecting this and not assuming that everyone who happens to have a German IP address speaks German, which is blatantly not true.
Yeah I agree. I also think western tech companies try not to display the fact that they could know a lot about you in their content recommendation algorithms as a sign of deliberate restraint.
In contrast go try out a Chinese social media such as Rednote, and the recommendation algorithm is creepily good at displaying that it knows what content you like. Yes it does feel creepy. But in this case I wanted it to know my interests.
I use ChatGPT with memories. The contextual memory is super hit or miss. It works like 40% of the time, but if context is actually important (and it is for a lot of what I talk about), I have to say "alright look up the memories for XYX for context" in like every message.
That works pretty well but me just constantly telling it to grep manually is a bit underwhelming as far as UX goes.
Maybe the answer to this is a proxy ‘you’. Like an llm trained to be like you, that you tote around and then ad companies and so on can interact with to find out whether their ad is something you give a damn about? Or the music services can bounce recommendations off?
Not sure I agree specifically with wanting companies to know more about me, but I do wish they made the most of what they know about me (at least in terms of providing me with something of value).
It's definitely a two-sided coin. I don't like the fact that they gather info about me, but I also wish that those few ads that do manage to get through were relevant.
What a naive take. The data collected by them is not there to serve you better with music or news!
You are not their client. And never been. And will never be.
Banks, credit unions, advertisers, insurance companies and the local and federal police. That's some of their clients.
I suspect it's just expensive to do a good job of it.
Amazon used to be better at targeted advertising than it is now. I used to browse for a few things on Amazon, then go do something else, content with the knowledge that they'll do the hard work of figuring out what is actually a good recommendation for it. For a few years, this strategy helped me buy tools, light fixtures, and other things I used around the farm.
I'm not sure when it stopped, but it definitely did.
Ultimately, it's probably just a lot more compute than it's worth. They can probably spend $1 in compute to drive $10 in sales with dumb ads or they can spend $2 to drive $15 with smart ads. It's a diminishing returns thing and they drew the line such that I get dumb ads. Maybe folks that buy more, higher priced items get better ads.
Claude is pretty good at this. I finally caved and gave it access to my calendar so I could run a weekly "check what's on nearby and suggest some activities" workflow. It looked at my past calendar activities, built a "taste profile" from that, and now regularly sends me tips for stuff I might want to go to. If I'm keen for something it adds events to the calendar, and optionally invites my partner. It's also surprisingly good at finding "wildcard" events which don't necessarily closely match things I've been to before, but are often interesting to me anyway.
Honestly, I've found it's easier and cheaper to just pay for content; instead of having my content peppered with annoying ads and then buying junk I don't want to support the content.
It's different with Meta advertising though. Back in the days Facebook was the only webpage that showed me relevant local targeted ads about things I care for, and nowadays it's the same on Instagram (that I also stopped using recently).
But for a while not too long ago, like for one year until last summer I regularly used Instagram, and almost all ads were directly targeted at my interests and they directed me to local events and concerts and products and what not.
They are in the business of selling you things, their positive signal is not "what makes your life better" but "what makes you buy more".
From my social circles I hear many people claiming how amazing are google ads and amazon/youtube recommendations. Which I have never understood, none of the online websites I use tend display anything relevant to me, when they do I bet its kind of random.
When I'm looking to purchase an item, it would be nice to get relevant ads for the kind of specs I'm looking for but like the blogger said, why do I care about getting ads for something I just purchased?
I've been in the same shoes as the Author. The ONLY piece of technology that I have successfully seen actually learn from my usage was Youtube. Just recently, youtube decided I needed a bunch of video game hate videos. But I don't usually want negativity like that, so I pressed not interested on two of them and now they're all gone. This has been the status quo for me for years.
Spotify, Instagram, all those others, it's like the button ENCOURAGES them.
I also stopped getting ads years ago, but it was the same story. Car Ads. I hate cars, I have made it as clear as possible on multiple occasions. Sidebar ads never learn. Amazon never learns. And I bet they keep trying to show me cars to this day.
> Netflix on Apple TV asks me to select a user every time I launch it. I've had this account for 15+ years, and never had more than one user on it. I always select the only option there is to select. There is no setting to disable this screen.
There are some remotes that can be programmed to automatically send a "select this" code I think, to automate having to do this every time (was researching something like this recently so just mentioning I think it exists, don't know a lot about it)
> When listening to personalized playlists on Spotify and YouTube Music I skip every single electronic track with vocals. I listen to a lot of techno and related genres, and the moment someone starts singing I skip the track. I've been doing it for years. None of these advanced algorithms have ever noticed this simple pattern.
Same here.
I think the songs get pushed by who pays to push, and most of those songs have vocals.