Hey, news websites? We need to have a serious talk about ads.
Just now, I saw a link to a story I was about to read (ironically, one about how best to monetize news on the internet), but attempting to click through to it yielded a massive malware warning because of the ad server
This is not at all uncommon. Moreso than any other sort of website, any given news site, especially the online mirrors for traditional newspapers, run ad banners that serve up malware that will completely ruin your computer if you allow it to load. When that doesn't happen, they still load up so many data tracking cookies that even with a high end gaming computer and a dedicated broadband connection they can put a noticeable strain on things or straight up crash a browser.
This has been an issue for close to 2 decades now, and the solution most people default to is to install some ad blocking plug-in, keeping these from ever loading. This is basically terrible for absolutely everyone involved, and it's a situation we only managed to get ourselves in because a handful of people, unfortunately including whoever you have managing your advertising, are grossly incompetent, morally bankrupt, and frankly, just stupid.
A brief history lesson-
Back in the days of print media, advertising was basically a black box. The publications set aside pages (or portions of pages) for ads, selling this space for fixed prices, to advertisers who could only send along images/blocks of text to place in it, and had absolutely no information whatsoever to go on about how their ads would perform other than the publication sharing their subscriber count/sales figures and demographics they appeal to.
This was, seemingly, terrible for anyone looking to place ads. Back in "the dark ages" they would have to just place an ad on blind faith that it would bring in more money from increased sales than it cost to place it, and no way to know if it worked than watching their numbers.
HOWEVER, and this cannot be stressed enough, THAT IS STILL THE ONLY WAY TO ACTUALLY GAUGE HOW EFFECTIVE AN AD CAMPAIGN ACTUALLY IS.
Anyone who has ever actually studied the psychology of advertising, or even stopped for a moment to consider their own spending habits, will tell you that effective advertising does not work in an immediate sense. Nobody sees, say, an ad for a new blender available at a particular store, immediately drops what they're doing, and rushes to that store to buy a blender.
What actually happens is they briefly see the ad, then days/weeks/months later, their blender breaks down, or they happen to be in that store, and the memory of that ad comes back to them that oh yeah, store is a good place to get a blender.
The ad works by just giving them a thing to remember later on.
Then there's internet ads.
Very few things on the internet got started by experts in related fields moving to the internet, especially back in the 90s. Mostly, a bunch of wheels were reinvented by a few stupid sheltered rich white kids, who got a heard start on internet adoption after programming classes they took in college because computers are simple and predictable and don't require you to deal with other human beings. As a demographic, such people consistently have absolutely no fucking idea what they are doing, but enough money and lack of competition to muddle their way to something that looks enough like success if you squint that you can con wealthy investors into pumping cash in that steadily inflates stock prices, attracting more investors, then bail out before that bubble bursts.
This is particularly true for advertising which those with long memories might recall was the actual specific variety of "internet" business behind the late '90s bubble that burst so badly. Particularly if you were writing on the internet at the time like I was.
As is common to idiot CS majors, these people got it into their heads that they would "improve" advertising by making as much data available to advertisers as humanly possible, allowing them to make better decisions. But basically all of the data they made available was really just totally useless noise.
First, of course, there's pageview impressions. Now, this one is kind of a necessary evil. Websites, generally, like to be freely available to read, where traditional print media has some actual purchase of a physical object involved even if it's at a severely reduced price. If you're selling your newspaper for 10 cents, that's not where your money really comes from, but it's where you get the exact number of people buying your paper from to tell your advertisers. You can't track "monthly sales" for a website, just access calls on the server and the IPs they go to, so that's what we ran with, and it's junk.
There is absolutely no standardization to how websites are formatted. Some have you constantly reading little snippets of text and hitting new links, some have one big long megapage, some split the difference with a floating ominpresent masthead/nav bar framing partial pages...
... and particularly in the early days of the internet, the ways in which people accessed pages varied wildly. Depending how your ISP charged, you might be miserly with your bandwidth, locally archiving a page and never actually loading it again, reloading it every time you had a few spare minutes where it was OK for your phone to be off the hook, or anywhere in between. Your browser might have defaulted to cached versions or not to the same effect, and of course people thought it was important to count unique IPs viewing a given page to make sure those numbers weren't being inflated by someone just repeatedly refreshing a page, but some ISPs would give you the same IP for life, others a new one every time you dialed up, and a lot of people were using library computers which shared one IP for dozens of people in a given day.
So the "highly accurate" data being given to advertisers was anything but, using it to project costs and sales next to the "murkier" (actually far more precise) numbers on print media would make it look like an unpredictable risk, and everyone was skeptical about it.
Quick aside, as I ramble here, I'm suddenly really curious about how billboard and marquis advertisements are priced. because I don't think advertisers get access to hard numbers on how many vehicles per day see one. Even if it's totally unknown though, it's gotta be consistent.
The next thing idiots with CS degrees did to destroy this market was start tracking click-throughs. Because while HTML is all about being able to click on or hover over linked text to skip around a large document, but web pages are largely designed by people who just use links to other documents because they're marginally easier to insert, we're constantly clicking through links to other web pages, and hey, you can turn an entire ad into one big link to whatever it's advertising, AND accurately track exactly how many people do that! Hard sales data! NO.
Trying to gauge how effective an ad is by how many people click on it to see what's being advertised is about as accurate as if you set up a billboard with a stack of order forms under it people could stop and fill out.
Again, nobody drops what they're doing to buy a thing whenthey first see an ad. When you first see the ad you are actively in the middle of doing something else. You're reading an article or watching a show or researching something.
You just register it to recall later.
Outside of really specialized edge cases, like how youtube spits out a list of recommended videos to watch next after you're done with something, I wouldn't be surprised if you told me nobody has ever actually intentionally clicked any ad on the internet. Why the hell would you, ever?
BUT, ads get sold on click-throughs.
And "unique" clicks.
This, even moreso than inaccurate exposure data, absolutely destroyed the perceived value of internet ads. Advertisers could directly see "how many sales they were getting," and it was basically zero.
In the late 90s it was not uncommon for a website that could easily boast a solid regular readership of 5000 people a day staring at prominent ad banners for hours at a stretch to be paid for those banners something like 2 hundreths of a cent per ad exposure, maybe a whole penny per click.
People tried to deal with this destruction of perceived value by
A- Really just cramming in a hell of a lot of ads all over the place while forcing readers to load more pages to see the same content.
and
B- Selling their pages to a handful of conglomerates, who would then present the combined readership numbers to potential advertisers, and have them all run the same shared complex ad banner code, rotating in a variety of different ads. And this brought in so many new problems.
First of course is the fact that those behind these conglomerates were complete scumbags, who took advertisers for a pretty decent chunk of change when first wrangling up their collections of sites, mostly embezzled that money and blew it on prostitutes and drugs (look up the relevant articles on your own time), and completely shafted everyone at the sites they owned when advertisers failed to actually see a matching uptick in sales when advertising with these conglomerates.
Also, we replaced a rather direct advertiser to publisher relationship with... a handful of parasitic silicon valley cokeheads throwing explicit animated porn gifs and malware on news sites and sites aimed at children...(whose actual staves had literally no ability to pull those, and weren't being paid).
So... that's how the '90s internet bubble burst, why internet ads pay basically nothing to this day, and why anyone who was writing back then is simultaneously dirt poor and nostalgic for print
You know who came out of that in a really good place though? The coked up scumbags with CS degrees selling slots in rotating ad banners to porn sites and malware distributors!
They never actually left they just gave up on buying out websites and switched over to renting space onnews sites.
Ad supported news sites are honestly relative latecomers to the internet all things considered, so this horrible broken mess was already in place, but switching to the internet from printis so convenient to everyone involved that they have the numbers needed to make this all sort of work. Kinda.
The people running the actual ad code though continue to be stupid, incompetent, chasing those direct click-throughs, and most importantly, being completely unscrupulous about what gets stuck in those banners.
This, then, leads us to this terrible and untenable arms race. In the face of internet ads underperforming based on the inaccurate metrics ad banner managers provide, grow increasingly elaborate, annoying, and invasive. Installing spyware on computers to try and work out better demographic info and SOME kind of real performance data, blaring loud music and full video, outright hijacking browsers with redirects to product pages, and increasingly becoming vectors for straight up malicious malware and porn/gambling sites which take what they can get.
People exposed to this horror do everything they can to keep this horrible and offensive garbage from violently interrupting them and measurably hurting their computer's performance.
News sites get more desperate as most of their readers are now essentially pirating their content to duck their advertisers, contracting ad servers that pay better in exchange for being even more invasive and annoying, and working on systems to detect and disable ad blockers.
Again, nobody wins. Nobody's actually seeing ads (or if they do, they're so horrible they resent what's being advertised), sites make peanuts, readers just deal with constant stress and aggrevation. Plus, you know, the actual content of these sites gets harder and less efficient to access.
View source on a plain text news story from a major paper. By all rights you should see the story and a couple formatting tags.
What you actually see is a few megs of illegible garbage spewed out by the ad code, which you're downloading every time you run a story, and which the news outlet is uploading. Lot of wasted data transfer cost there on top of everything else.
Plus seriously, news sites, you are currently making all of your money by selling opportunities to utterly malicious entities to actively harm your readers with invasive malware, and that's really not a good position to be in.
But the fix for all this is actually super simple!
Specifically, you just transition to a simpler ad model. The less complicated the better, really. Cut back on go-betweens as much as you can. Cut back on providing metadata as much as you can. Keep ads as mechanically simple as you can, with the lowest file size impact you can.
The ideal goal would be to
just have a plain text article BUY MOXIE
with a little static jpg or a
piece of text off to one side,
just like you'd see on a print CONNER'S LAWN
page. MAINTENACE
This isn't me just talking theory, either. This is actually tested. Everyone who had an internet based business in the 90s who's still alive worked this out after getting burned by the terrible predatory ad servers you're using. You can see it with Google's little text ads you barely register, or those Project Wonderful ads that webcomics have rather strongly embraced. Simple, unobtrusive. I've placed both, getting comparatively little direct data. And running them, they generally pay better than the gross flash banners to boot.
