From Charlie (NYT; Atlantic) Warzel's Galaxy Brain substack, September 23, 2021:
This is the longest Q&A I’ve ever done here on Galaxy Brain. There’s quite a bit, here. But I think it’s worth the read. Let me explain why. My first reporting job was at the advertising trade magazine, Adweek. When I started, the ad industry was just beginning a wild, digital transformation. This thing called programmatic advertising was taking off. Ads were sold on these exchanges in these split-second auctions and it felt like dozens of companies were springing up every second. I didn’t fully understand it, but it seemed not all that nefarious at the time.
But the more I started reporting on digital privacy issues, the more it became extremely clear that this huge, opaque online advertising ecosystem was a lucrative blackbox, powered by our personal data. It grew faster than anyone could regulate it, so it was barely regulated. It was (is) ripe with billions of dollars of fraud per year ($52 billion in 2020). And because the system is so hard to understand, it’s ripe for bad actors to use it to pollute the internet with garbage ads. The online ad ecosystem is helping to fund most every publication — including those that spew hate or misinformation. The online ads game is at the center of Google and Facebook’s dominance, too. Most everything bad about the internet (and a lot of good things too) has at least a little bit to do with this ecosystem.
What you’ll realize the moment you start reporting on this world is that it is convoluted and extremely difficult/technical to parse. To do a decent job, you have to immerse yourself in the weeds. Shoshana Wodinksy (a former Adweek alum, too) over at Gizmodo is arguably the reporter that is deepest in the weeds on this subject. Wodinksy isn’t afraid to comb through pages of marketing pitch decks and obscure developer reports or an apps code. And it shows. Her explainers help me understand what’s actually going on in the bowels of the internet and where the money comes from. I caught up with her last week to talk about online advertising, which I sometimes think of as the internet’s original sin.
Hello. Let’s talk about adtech. How’d you get started on this beat?
So, yeah, I never really planned to write about adtech.
How’d you get sucked into it?
I graduated journalism school at the end of 2018 and I really wanted to write about tech stuff. The folks at Adweek reached out and said, ‘this will sound crazy but we’re hiring for an adtech reporter and think you’d be good.’ I didn’t know hardly anything about adtech. My soon-to-be editor was interviewing me and explained adtech via how it related to the Cambridge Analytica scandal, which was really fresh in everyone’s mind. He was basically like, ‘you know, there are thousands of companies doing very similar things as Cambridge and getting away with it and getting away with a lot of other stunning stuff, too.’
Trade publications are such a great education.
At the trades you start learning the basics, especially the grammar and industry vocabulary. And this adtech shit was mind-numbing. So I spent hours late at night, just reading and pouring over developer documents from unknown ad exchanges. I got extremely lucky because I started out working with Adweek’s other adtech reporter, who was based in the UK and he was so jaded the way great, grizzled journalists are. He was like, ‘never take these companies at their word.’
I started poking around with this in mind and a lot of the company promises fell apart really quickly. What better way to hold power to account than to write about ad fraud and focusing on this mind-numbing weeds-y tech stuff. When it came to small, but meaningful privacy issues, I ended up learning about a a ton of stuff the ad platforms didn’t tell the public. Trade audiences already knew that this stuff was happening but the broader audience didn’t.
This really tracks with what I’ve found over the last decade of learning how to report. I think it’s this way on most beats but I find it especially true in covering privacy and data issues. A lot of this stuff is like, purposefully dense and dull. And the most interesting stuff is always the most impenetrable shit. You basically need almost like an advanced degree to understand how your data moves across the internet.
This is what strikes me about your work. This one piece you wrote in 2020 -- How Google Ruined the Internet (According to Texas) -- has a catchy title but is ultimately an incredibly detailed explainer on the mundane workings of the online ad industry. You start with a primer on how ad servers and exchanges work (basically: detailed automated auctions that happen in less than a second that are fueled by the data companies collect on you based off your browsing history and other collection methods) and move into the way that Google’s dominance in the ad industry forced all kinds of scummy things to happen, including some colluding with Facebook.
Ultimately, you show how wildly boring tech stuff like “header bidding” leads to this race to the bottom that hurts competitors first, then advertisers, then anyone who publishes things online. And the internet turns to this kind of garbage-y, pop-up-y, clickbait-y surveillance hellscape. I guess what I’m saying is that, if you don’t understand the exchange minutiae, you’re not going to really know why the internet is decaying under the weight of these companies.
Back in January there was this controversy over WhatsApp’s privacy policy. Just a ton of confusion and it was clear that both all the people freaking out over the change as well as Facebook’s comms department didn’t really know what they were talking about. For my part, I’d long been familiar with the quirky parts of WhatsApp’s privacy policy and its business model and so I wrote a 2,547 word explainer on it. I screenshotted some of the code that shows what the data collection looks like. It allowed me to show readers that, while the policy update sounded awful, it was essentially just Facebook/WhatsApp saying what they’ve always been doing out loud.
And that, to me, was a good example of demonstrating that, hey, this is a company that exists to make money and ultimately any data collection happening on you is going to happen for profit. So let's talk about who's making money here and where that money is going because if you understand that, then you'll understand how we can fix these things for the better in the future. As you can see, I’m very fun at parties....
....MUCH MORE
