3 ms·
> If the point of twitter is to make money, then it doesn't matter whether someone is logged in or not: ads in front of eyeballs is what's important. To serve
by jdminhbg 3y ago
> If the point of twitter is to make money, then it doesn't matter whether someone is logged in or not: ads in front of eyeballs is what's important.
To serve you ads that have any hope of making money, they need to know which eyeballs they're going in front of.
- burkaman 3y agoNo they don't. People sold ads profitably for hundreds of years without any sort of surveillance, and many people continue to do so today. If it were impossible to make money just selling ads (I wish it were) then billboards and bus stop posters wouldn't exist. You can even make money on the internet this way: www.ethicalads.io. Twitter also has an enormous advantage: even if they don't know anything about you, they know a whole bunch about whoever posted the tweets you're reading. They could do a lot of targeting by just looking at the content you're viewing.
- jdminhbg 3y ago> People sold ads profitably for hundreds of years without any sort of surveillance, and many people continue to do so today. If it were impossible to make money just selling ads (I wish it were) then billboards and bus stop posters Billboards and bus stop posters are cheaper to run than Twitter.com is, so making much less money is ok. The whole reason Twitter has always been unprofitable but Facebook/Meta is a trillion-dollar public company is that FB figured out how profitably identify users and serve them ads and Twitter didn't.
- burkaman 3y agoI agree with you, it is unquestionably more profitable to spy on people and then sell them ads. My point is that you can still make money without the spying part, and there is no reason for us to accept that the two always have to be bundled. If you can't make a profit without spying on people then your business model sucks. The fact that you might be able to make more money by being less ethical is not very interesting, that's true of almost anyone.
- abracadaniel 3y agoNetwork television is enormously expensive and entirely paid by ads without knowledge of the viewers. Instead they make assumptions based on surveys of demographics and manage to be very profitable. I wonder if you could make a comparison of ratings of a TV show to active users of a website. For example: is Twitter the equivalent of daytime TV, or multiple series-finale-of-Friends per day. *edit: it looks like they are the equivalent of 0.8 Friends finale / day. 41m daily active users vs 51m viewers of the Friends finale.
- MisterBastahrd 3y agoCookies do this for them.
- rchaud 3y agoIt absolutely matters if someone is logged in. 3rd party cookie blocking on Safari, Chrome and Firefox has crushed the ability for advertisers to do remarketing with any level of precision. That was where you didn't need the user to be logged in, you just needed to match their cookie ID on Twitter.com with the cookie ID passed on by Site X, which the user had visited previously. Nowadays, Advertisers upload customer lists with email, phone numbers and other identifiers into Twitter Ads for ad targeting. Twitter shows those advertisers' ads to logged-in users that have the same email/phone # or other identifier. https://business.twitter.com/en/help/campaign-setup/campaign-targeting/custom-audiences/lists.html https://business.twitter.com/en/help/campaign-setup/campaign... Anything else is effectively untargeted and useless.
- autoexec 3y agoThey already have the only information they really need, which is what you've requested to look at. If I'm following someone who tweets about X you can safely guess that I'm interested in X and target me with ads related to that without ever knowing my identity.
- jdminhbg 3y agoThere may be a reason there's no social media network that does this.
- autoexec 3y agoTo some extent they very often do use the type of content you consume for various things, but they don't use that data exclusively because they can make far more money by collecting, leveraging, and/or selling every scrap of personal data they can get their hands on and money is really all they care about.
- foobarchu 3y agoIt may be that it wasn't feasible on large scale to match content to ads. I feel like recent advanced in machine learning and ai could change the game there. What were describing is essentially how television ads worked for decades, isn't it? You know the content, so you can identify the demographics and advertise properly. Podcast and YouTube sponsor deals often work that way too (not always, but that's on the advertiser).