4 ms·
>I have been running a server for a year or two. Paying for that myself. I get a wage every month from the job, so I can pay for a server. Theoretically I could
by ApolloFortyNine 5y ago
>I have been running a server for a year or two. Paying for that myself. I get a wage every month from the job, so I can pay for a server. Theoretically I could run lots of services on that server and still only pay the same amount every month. For dedicated people in IT sector wages are often good and they can afford to run a few things out of their own pocket. My guess is many people would do that. Not every website needs to be "financing itself". It is not always about the money. Some people simply want to make a nice thing and are OK with paying for it.
This does not scale. At some point, you need to make money somewhere.
>Besides all of that, ad business is often make-believe by the big players, giving wrong impression of how much an ad actually helps your business and improper conclusion drawing from statistics by marketing departments, instead of data analysts.
>How many marketing departments are capable of running a proper A-B-test?
Again your just digging deeper, further calling out a trillion dollar business for being wrong. Besides that you would likely need thousands of sources to accurately back up such a claim (since there's people paid much more than you, with access to many more resources than you have, have decided this is worth it). You are literally calling out entire departments that likely have a payroll 1000x your salary.
- zelphirkalt 5y agoEven greater their incentive to keep playing the game. Many more wages depend on that than my own. I am not necessarily calling for them being wrong, but for them playing with the numbers and presenting them in a way, that makes many people believe. They have credible deniability as well, because their tracking script didn't even run on browsers, which blocked it, so they couldn't know about those "edge cases". Only that those "edge cases" can make a significant portion of the total visitors of that website. In the end they profit from telling everyone, that they must use GA (or similar tool) to track what is going on on the website and most marketing people will happily jump on that train, because it gives them any kind of data, which they can use to justify things, even if that data is only half the story and cannot be relied upon to give a true picture. "The data tells us so!" makes the job much easier, unfortunately often at the cost of user privacy. And so the make believe, that you must track your users with third party trackers continues and propagates. Then on the development side of things, developers or their higher ups eschew the work needed to implement first party tracking. They want that cake at no cost. Without strong ehics, the website of such an organization is doomed to disrespect the privacy of its visitors.