It would add a bit of technical complexity for site owners, but that seems manageable, particularly for b2b relationships.
Obviously, I hope this doesn't happen, but it seems like an obvious strategy and I don't see the flaw.
I also would like to know why there is so much resistance to this, which was the original model of ads on websites ...
rsync.net stopped advertising, in all venues, about two years ago - mainly because the overlap between "people smart enough to use rsync.net" and "people who don't use an adblocker" is basically zero. Nobody who cares about our product ever saw our ads.
But, of course, we still have some interest in advertising our product and, to that end, I have approached several websites and offered very good money to just insert two lines of plain text on their HTML page. No "network", no code blob, nothing interactive ... no picture ... just an extra line of text, with a bit of it href'd for a link.
Huge pushback on that. No interest. "Impossible".
I really don't understand the responses I've gotten ...
"What happened to text ads? The text ad facility allowed readers to place simple, text-oriented ads on the site. Use of this facility had been dropping over time; when we realized that nobody had bought an ad in over six months, we decided to remove the feature."
LWN would still sell you banner advertising, but they don't do text any more. As with other features that were killed because nobody used them, the interest of a single small buyer won't bring them back because it doesn't make any economic sense.
I figured someone at their organization had the wherewithal to open a regular file in vi and paste in two lines of HTML ... in exchange for money ...
Nope.
I once had the idea of delivering ads by dynamically injecting them into a video stream. As in, when you load a youtube video it prepends the ad into the same video stream as the main content. Ideally the transition from ad to main content is between key frames so that the ad-blocking software can't fully strip out the ad without screwing with the stream itself.
My objection to advertising on the net is primarily the tracking that comes with it. I don't use adblockers specifically, but I do block things like Javascript and tracking pixels as much as possible.
Given the nature of modern internet advertising, I would assume that the tracking would be happening regardless of whether the ads were being relayed through the host website or not, and it wouldn't change my security stance.
It would make me much more suspicious of the website, though, as I'd wonder if the site were sharing its log and other data with the ad network. This is pure speculation, but I wonder if websites might be nervous about bringing that cloud of suspicion over them.
I know that if I were asked to include a barebones ad (I assume that it's text-only and doesn't link to an image you host) like you describe on my websites, I'd decline.
I think I'm not describing the proposition I made - it was, literally, paste this line into a page:
<a href="https://rsync.net">rsync.net</a> - Cloud Storage for Offsite Backups
No tracking. No engagement. No stats. No performance tracking. Nothing. Open up a regular file in 'vi' and paste that line into it.
Impossible.
The cost of potentially blocked by ad blockers is finite (A percentage of total revenue), but the cost of ad-fraud is not bounded.
Isn't the solution to that problem a flat rate fee (similar to how advertisements on tv, newspapers and magazines work)?
Instead of a pay-per-click it could be a simple $X dollars and your ad will be visible for Y days/weeks.
The current solution is effectively a flat rate as far as an ad campaign is concerned: impressions/$
Needless to say, this could be inconvenient for the adwords-make-me-five-bucks-a-month scale sites. It'd work out OK for the New York Times-es of the world though.
But even with the added complexity of regularly pulling in the public suffix list, the problems keep going: e.g., facebook.com's scripts are all served from static.xx.fbcdn.net.
The main problem I can see is that you would have to choose between CDN caching and targeted ads. If the video was uniquely generated for each user with new targeted ads, it couldn't be cached.
I know that some podcasts use a technique like this, modifying the audio stream at download time to insert recent ads, even if the podcast being downloaded is old.
I suppose it's much more a hassle to get rid of the ads from the app.
There's no API needed, all they require is for the website to set up a CNAME record. They're usually random and rotated often.
If you're familiar with the current arms race in ad fraud, there's a coming arms race for ad blocking.
Wouldn't this do the exact opposite? Ads that did this wouldn't be able to track clients across different domains, even for clients that allow third-party cookies.
But I think the main issue is the jump in technical complexity, which is more than "a bit"—many, many websites are built by non-coders who can copy and paste a bit of HTML & CSS, perhaps using a CMS or blogging tool like WordPress or Moveable Type, and who can paste in a script tag, but no way could they write even the 3 lines of PHP it would take to reverse proxy for an ad script. Maybe if this approach became popular enough, there would be a standard WordPress plugin or something, but what adtech company would want to be the first to require it?
As well, obviously the owner of the website becomes directly responsible for malware served through an ad network.
Finally, advertisers prefer third-party to fight ad fraud (by the site owner).