Hacker News Reader

Comment by blakesterz | Hacker News Reader

I found the actual report this is based on even more interesting:

https://app.urlgeni.us/blog/new-research-across-200-ios-apps...

Just a few of the highlights:

Magazine apps had the highest number of total network contacts (28), and the highest percentage of third party domain contacts (93%)

Social apps, followed by Games apps, made the fewest number of network contacts, 6 and 7 respectively.

Apps making the most number network contacts included iHeartRadio (56), Wall Street Journal (48), ESPN (42), Popeyes (42), and WattPad (36)

View on HN

Agreed, the underlying was more interested than the summary. And also seems to be a case of measuring what's conveniently measurable.

The fact that {product} has only 1 connection to {product first party domain} doesn't say a lot about anything, given they could be internally proxying to and from who knows how many partners?

It'd be more helpful to at least see total traffic per domain.

It made me curious that the highest "first-party ratio" companies tended to be tech companies capable of realizing their own architectures (Amazon/Google/Apple).

Yes, third-party direct connections from apps is useful information in some ways and I'd prefer to keep it down rather than spread things around widely.

But what people are really worried about with Facebook or TikTok or whoever building and aggregating information about them, doesn't require talking to a third party at all, and if you clamp down on these direct connections everything could just pass through the "first party" and through to whoever they wish on the server side where you can't see the connections anyway.

Revenue streams.

These network contacts provide $$$ to those companies. They're low effort, easy revenue for them. Privacy for customers be d*mned.

I once met a senior-level dev who worked for a company, they wouldn't have their own app on his phone, due to those invasive practices tracking his behaviour.

I thought the most curious was the ratio of third:first party domains for magazines and news.

It suggests they control very little of their technical, revenue-generating infrastructure.

Which is interesting when the narrative is "The internet is destroying journalism." But another way to phrase it might be "All of these companies drastically underinvested in web/mobile, and continue to do so." (Although the NYT seems to have a higher ratio than most, but I believe they divorced themselves from Google and brought a lot back in-house?)

It's a matter of scale in relation to your audience and data. Your local news site with users in the tens of thousands, is unlikely to have the capital to invest in geographic and demographic tracking necessary to provide a robust ads business. Easier to plug into a number of providers who already have that data about their visitors.

These sites have to do that to remain competitive.

Also, the report makes the title seem very biased. Home Depot, Reuters, WSJ are all using more 3th party trackers.

I think who they are sharing with (potentially) matters more than the number of orgs they are sharing with.

Reply on news.ycombinator.com