Hacker News Reader

Comment by captainmuon | Hacker News Reader

It seems like the tracking protection looks for specific parameter names. Won't people just start using randomized names?

Already now I have to rename my stats script from 'track' to 'res', otherwise I only see about a quarter of users. (I'm just using it to count how many views each page has, not for tracking.)

The endgame is to move all ad and tracking code to 'first-party', i.e. the website delivers the code in its source. (Un?)fortunately that is not happening quickly since advertizers distrust the publishers.

View on HN

> Won't people just start using randomized names?

No, the UTM parameters have been standard for almost twenty years and everything is built to follow them. The whole point of UTM parameters is to follow a user from site A to site B (on site A you can just track what links people click), so randomizing it won't work. It needs to be coordinated.

The target site has access to all the query parameters. If they wanted to they could send all of them to Google/whoever who then looks up if any of those parameters are in their database

That's an API someone has to write, and Google just burned down Analytics for no particular reason. It won't happen. The UTM parameters were a Schelling point. They won't be replaced because once you try to coordinate something else, Apple will block that too, so no one is going to bother trying. Google will be happy to just keep the data to themselves instead of sharing with others.

>The endgame is to move all ad and tracking code to 'first-party', i.e. the website delivers the code in its source.

I don't know why this hasn't been happening. I can think of a thousand ways of making sure that users see my ads. The fact that you can block ads via /etc/hosts is... pathetic.

The original sin of the web is that publishers let third parties take over advertising, then surprise, surprise, the third parties kept all the money to themselves and choked the publishers to death.

If you just want to count page views you can use your HTTP server logs, no need to have Javascript.

There's a reason we don't do that anymore and it's that there's too much noise and it's easier to filter it out if you aggregate the data somewhere (bot traffic, scraping etc.). Sometimes you also don't really have access to HTTP server lots depending on where you deploy your frontend.

I don't think that would play well with cdn caching.

I used to do that almost 20 years ago with a little script.

This time I looked into my log file, said nope, too complicated, and just installed a WordPress plugin (or Matomo in another case).

> The endgame is to move all ad and tracking code to 'first-party', i.e. the website delivers the code in its source. (Un?)fortunately that is not happening quickly since advertizers distrust the publishers.

The other issue with this is that advertisers use Google/TradeDesk/Etc because it gives them access to a ton of supply (publishers) with one interface to handle all of their reporting/metrics. If the publishers own the tracking/targeting/reporting there will be 1000 different implementations. Which is probably good for the end user and publisher but bad for the advertisers. It will be interesting to see where adtech goes in the next 5 years.

Reply on news.ycombinator.com