Zero per cent of app users are running ad blockers
Not to take away from his point, but… Pi-Hole.
I use AdAway to modify the host file on my Android phone. I literally can’t load ad content; sometimes it gets in the way since the ad is the easiest way to get to what I want to click occasionally, lol.
Also, the Vanced app can modify apps to eliminate ads. Or X-Manager for Spotify does a similar thing.
Then there are alternative front-end apps, like NewPipe/LibreTube for YouTube, Xtra for Twitch, and others.
Or there is just straight piracy to avoid the ads, but in their own apps. Stremio + Torrentio + a Debrid service is basically a pirate Netflix app with full control of your stream quality.
Regardless, I think his point mostly stands; it’s a very small percentage of users modifying apps to avoid ads or telemetry.
Yeah, <1% would be the more technically accurate answer.
I’m still not sure why that works, honestly. If I was going to be evil I’d load the ad data via the same packets as the app payload. It wouldn’t be hard and there’s no way I’m the first to think of this.
The vast majority of mobile ads are served from google or apple, which likely is not where the rest if the app data is served from. It makes it particularly easy to block mobile ads from adMob or whatever the apple equivalent is called.
You don’t even need pihole, there are VPN-based apps that screen and block ad domains.
If too many people were blocking the ad servers, apps could tunnel the ads through a single connection to the app… but that would only work for ads served by the same platform as the app (like Meta ads in Meta apps, Google ads in Google apps, etc.), with 3rd party ad networks not trusting random app developers’ tracking and engagement data to send them a payout… so blocking the ad servers is likely to keep working for those, and for the larger platforms as long as they don’t see much value in spending resources to counter it.
with 3rd party ad networks not trusting random app developers’ tracking and engagement data to send them a payout
Ah, that’s a good point! I hadn’t thought of that.
There’s some approaches you could try there, but at some point I could actually be helping somebody, so I’ll shut up.
And, ime, a lot of corporations are serving content through third party (or at least non-native) servers, which means that any blocker which touches any of those servers breaks content completely. I’ve experienced major Travel, banking, and retail sites which simply don’t work unless most blacklisted sites are allowed. That means either turning blocking off for that main site entirely, or spending an hour testing every one of their 30 off-site connections to see which ones break. I don’t have that kind of bullshit time, and the rest of my family don’t have the patience or skill to do that troubleshooting. PiHole turned out to be multiple hours a week of frustration so I gave up - I already have a full time job and full slate of hobbies. In-browser blockers are, at least, easier to toggle on and off.
I agree that pi-hole is an option here, but yeah, the reality is that most daily users don’t even know what it is. At least, not yet.
Adblock Plus and it’s betters became ubiquitous in large part because they were so incredibly easy to install. As easy as gramma accidentally installing yet another yahoo toolbar on accident. Like, too easy.
Pi-Hole isn’t hard to install, and there are some fantastic guides to help users get it running with essentially zero prior knowledge. But in my opinion, I think until it gets closer to “push-button” easiness, pi-hole and systems like it will really be limited to the <5% of users motivated enough to go through the steps, who aren’t mortified of logging into their router’s admin page. I want us to get there faster, and we’re a hell of a lot closer than 10 or even 5 years ago. But we’re not quite there yet.
Edit-typo
Closest thing to “push-button”, are probably VPNs with a built-in ad blocker… which might as well be a pihole. Not sure what % of people use those.
Another way is through DNS (eg. noads.libredns.gr).
