You're viewing a single thread.
I've said it here before and I will say it again: Facebook and Google's entire approach to ads from data is based on an incorrect assumption of using enough data to build a profile on a person to predict what they are going to be interested in, when if you stop and think about it, it doesn't make any sense: people's needs will always change with circumstances at the time, ex. no amount of data in the world will help you predict whether I'll want a burger or Thai food for lunch tomorrow even if you do the digital equivalent of creeping on me outside my window and digging through my trash can. If you want to know what people want to buy, why not ask them?
Unfortunately, they've deluded themselves for more than a decade with the lie, so now the rest of the world also think that internet ads by them actually works.
I'm not going to rag too hard on Facebook here because most thing has been said by others already, but I will say that since literally every single one of their social media product took the same approach of maximize data to sell things, why is it that they are all losing people and their most successful place at selling things is actually Facebook Marketplace?
I agree. I hate the idea of Meta or anybody having that much data on me in one place. However - - even Amazon, which supposedly has some of the most sophisticated purchase data and analytics out there doesn't seem to be able to do anything with it. I mean, I buy a blender there and the next thing they recommend to me is... A blender. That only stops until I buy something else which they then continue to recommend more of, as if I only want multiples of the exact things I've already purchased. The recommendation engine is the whole point of having my data to begin with so but they seem unable to recommend anything new to me at all.
It's like they have all this information written in front of them but they can't read.
Amazon's recommendation engine feels like it is still designed for media, even though they sell everything now. Buying a blender doesn't mean you want more blenders, but buying a science fiction novel probably means you want more science fiction novels.
While I hate the shady data harvesting practices of companies like Meta, I do want to play devil's advocate here, as far as the value of data goes, if only for the sake of me understanding the shortcomings of it better. If a company were to dig through your trash can to get an idea of what you want to eat, so to speak, they'd probably find data on a history of foods you have eaten, if you've been interested in burgers, or any other foods you've been interested in. Or if you've been an adventurous eater in general or if you prefer to stick to variants of the same stuff you normally eat.
It may not give you a foolproof way of knowing what your next actions will be, but wouldn't it give a company an educated guess, at the very least? Enough to improve the chances of targeted ads being more effective, as opposed to missing altogether.
If catching the user's interest is a dice roll, then wouldn't the data at least improve the odds of rolling a number you want?
Ok, think of it this way, if your friend ask you for a local restaraunt recommendation, would you dig through their trash to find their leftover to figure out what kind of food they like, or would you just ask them what they usually like to eat?
That's why people went to forums like reddit for help with everything, because people helping people with volunteered info always work better than being a creep.
Oh yeah, I agree it's more effective, by far. I imagine that's why Google has Opinion Rewards and other companies use surveys to directly understand the needs of their customers. Though getting people in mass to volunteer that info, especially without some small profit incentive like Opinion Rewards might be tricky.
At least in the realm of targeted advertising, the closest example to user input would be when you set up an account, you're prompted to select your interests. Like with Windows 10, when you're setting it up for the first time, you'll be asked about your interests, which Microsoft will then use to send ads and news pertaining to those categories.
But yeah, I see what you mean
Ads are so backwards in general.
Why would I trust a brand just because they inconvenience me for 30 seconds before I can watch a YouTube video I'm actually interested in? Or the branding that interrupts me reading an article with a "trendy and hip", horribly made auto-play clip that bolsters an audio volume level so high, it almost deafens me?
I almost actively avoid products that I see advertisement for on a daily basis because they piss me off so much.
I get wanting to promote your product and it's not the company's fault that many places aggressively spam ads in your face to nickle and dime their user base.
I really feel like ad revenue should not be based on how many ads you show, but how much return the advertiser gets. Google's ad service has a return of about 4%, which means for every 100 bucks a company pays to Google to advertise a product, they can expect to recoup 4. I dunno how that is even worth it.
And yet, it continues... I think they're thinking that brand recognition is all that matters and that eventually you'll come around and get Geico or Use some product that has been assaulting you for months simply because you've seen it enough and it's stuck in your mind. Honestly, I would rather have a customer associate good thoughts with my brand, but no one seems to ascribe to that thought anymore....
I agree with what you say as one aspect of it. But there could be insidious uses of your data. What if your insurance company pays a couple of raccoons to sort through your trash and find out you eat fatty burgers and fries for most meals? When the raccoons get back with that report, what is to prevent your insurance company from raising your premium since you are at risk (according to them) of a heart attack?
They can't identify particular and specific interests, but they definitely can predict trends on consumer behaviour. I don't think Google wants your data to predict when you will want Thai. That's preposterous.
They are reaping billions on ad revenue, it definitely works, it won't forever, but it does at the moment.
They do their best to get your money. They monitor your past behavior to know what you're likely to listen to in the future. For example, if you eat basic burger/fry foods, they'll send you more ads for those types of places, but if you tend towards Mediterranean, then they will send you those places instead. As another example, dog food ads aren't worth sending to people who don't have dogs.
What I worry about is Alexa and the other voice boxes. Every time we talk about getting food around it, I open my phone and have ads for the places we were talking about. That's literally creeping on someone.
The issue is that while even if they build a perfect profile, they are still limited by which advertiser is paying more.
Google might know you don’t like cheeseburgers, but CheeseBurger Inc. just paid $200 million to have ads run, so you’re getting cheeseburger ads so that Google can’t report to the client that they’ve served ads.
no amount of data in the world will help you predict whether I’ll want a burger or Thai food for lunch tomorrow
they TELL YOU want to want. that's why you see ads for fat juicy burgers. will everyone who sees those ads go out and get a burger? no. but a few do.