The problem is with the title part of the url, since the title contains the word "ads".
I did an experiment: If you remove the words after "-google" in the url, it still works. You could probably remove the entire title of the url and it will still work by article id... So here's a link that will work with Adguard enabled:
Wow, I'm running a huge set of Privoxy rules I've worked on for the better part of two decades which tends to err on the side of blocking, and yet it doesn't get tripped by this positioning of "ads", because it's in such an innocuous position.
Try Microsoft ads, they aren’t backstopped by Google in any way.
Regardless of ownership, you’ll run into same problems almost everywhere if you publish articles critical of your advertisers. That’s nothing to do with Google and just the market working as intended/expected.
What's the use case for using both Plausible and Matomo at the same time? Just pseudo-"A/B" testing for which one is a better fit or are there major feature differences where using both is needed?
Recently there have been a few companies formally distancing themselves from Google, and doing so with a press release. Is there a reason why this is happening now?
To my knowledge, Google's basic nature has been known for quite a while, and people who care have lined up on different sides depending on their ideologies and self-interests. What has changed in the last 6 months that makes so many companies change their opinions about working with them? Is it just that public opinion has crossed some tipping point where it makes sense to gamble on a temporary revenue hit in order to be on the "right side"? That seems unlikely, just because I don't know a lot of people who weren't skeptical of Google a year ago who are now.
Is it related to FLoC? That seems like a wonky issue that's under most people's radar. Is it the EU antitrust hearings? Also seems unlikely.
Google's pushing advertising clients away. Techdirt started this path by at least August when they reported on Google flagging their pages (third link in this story), and it took time for them to shop competitors who are really resellers.
I would have guessed it would be most of their readers blocking easy-to-block ads, and the low payout, but it seems capricious policy enforcement is happening across Google and Youtube especially.
I'd dare say that the growth of DuckDuckGo, Apple, and a number of different VPN services are proving that consumers are highly interested in privacy and don't want to be tracked everywhere. This is also reenergizing the subscription vs ad revenue debate, particularly for big companies that have a vested interest, or, 30%, for pushing subscriptions.
There are several group dynamics I see in this. The one I think they want you to focus on is that typically nobody defects until someone high profile does. Then the floodgates open and everyone defects. Somebody has to go first, and sometimes several somebodies if the tribe is big enough.
More cynical dynamic is virtue signaling, which can pretend to be the former.
I think its as simple as this: people are fed up with FAANG; it seems like Congress might actually get off their asses and do something, and a few people smell blood-in-the-water.
You know, the sort of thing that could potentially kick off a frenzy… as unlikely as that seems.
In order to understand this, we need to explain the "Adpocalypse" - a thing that happened to YouTubers and, more generally, online content publishers that sell ads through Google. More generally we need to talk about the changes that advertisers have made with respect to how they think about brand association.
The way advertising companies used to treat online advertising a decade ago was that it didn't matter where your ads were placed, as long as they were going in front of people who were interested in them. That's why you had large ad exchanges with third-party tracking tools to build up behavioral profiles, because it was the best approach for linking advertisers and publishers. Publishers generally didn't care about who was advertising on their pages and advertisers generally didn't care about what pages their ads wound up on. This is, of course, a generality; there's always been picky advertisers and/or publishers, but the core point remains: the business worked because inventory was fungible.
And then advertisers realized that their ads were showing up on videos of Klu Klux Klan rallies and other far-right nonsense.
Well, they didn't so much realize, as much as they were publicly embarrassed when journalists started writing news stories about it. Consequently, you saw multiple waves of brands making a show of pulling all their advertising until exchanges implemented better controls for brand association and advertiser safety. Of course, this means that exchanges wound up over-enforcing ad safety, constantly harassing publishers that might be on the line. Journalistic outfits were unusually impacted, because their content is rarely advertiser friendly, even though they're not the right-wing nutters that started this mess.
This problem isn't related to trying to keep right-wingers out of ad money, though. There's a sort of converse Adpocalypse that happened to publishers, called malvertising. This can take many forms - from fake popups for scam tech support companies, to just outright dropping a zero-day exploit on your victim. This causes reputational problems for publishers, since safe-browing services will flag your site (basically making it unusable without lots of scary warnings, and heavily punishing it in search engines) if it starts dropping drive-by downloads on users.
So, advertisers are unhappy because of bad publishers, and publishers are unhappy because of bad advertisers. What about the users? Well, they're installing ad blockers in record numbers because they don't want to get hacked by malicious ads. Also, many of them aren't actual users; the online advertising business has had click fraud problems way before malvertising or brands getting cold feet over what content they're paired with. That's actually part and parcel of why online advertising is so tracking-heavy; it lets them detect the difference between an actual reader and someone just clicking on their own ads all day to defraud advertisers. Guess what? Users don't like ad tracking all that much either (especially when exchanges are so directly in-your-face about it.)
In any fungible market, the presence of "bad" commodities drives out "good" ones. What you have now is a toxic market, one that Google more or less owns outright because everyone else has been driven out of business or acquired. Google isn't the cause of the toxicity, but they benefit from it regardless, since Search display advertising will remain brand- and user-safe in ways publisher advertising isn't. In this new market of TV-like, non-fungible advertising; Google is no longer a business partner. They are a hostile, dogged competitor paying them ever-shrinking table scraps.
I’m sure it’s not the only factor, but loading JavaScript from external domains (like from Google) changed from being good for caching – i.e. faster web pages – to being meaningless and usually slower:
Two steps in the right direction, but this one page still sends requests to soundcloud.com, rawgit.com, cloudflare.com, googleapis.com, amazonaws.com, fontawesome.com, sndcdn.com, stackcommerce.com, stackassets.com, stacksocial.com, and gravatar.com.
Their "self-hosted" Plausible content is pulled from tics.techdirect.com, which resolves to 46.101.161.209 and shares a reverse DNS with custom.plausible.io. Can't tell who runs the web server on the other end of this, so it may well be self-hosted.
The site also dropped a first-party one-year cookie, and let a third party drop a week-long cookie.
My browser is set to send Do Not Track requests, so my explicit wishes are still being flagrantly ignored.
There is still a lot of third-party tracking going on here, as well as execution of remote code under the control of third parties. It's remains a privacy and security clusterfuck but it's better than it was.
Wait, they're still serving things via rawgit? They started shutting down in 2018! That service has been deprecated for three years. That sounds like an attack that is just waiting to happen.
> GitHub repositories that served content through RawGit within the last month will continue to be served until at least October of 2019. URLs for other repositories are no longer being served.
> If you're currently using RawGit, please stop using it as soon as you can.
Fire and forget mentality. Wasn't long ago that some advert video hosting site's domain was taken into new ownership to host porn. It must take a lot of effort to integrate all this third-party garbage into a site. If only a site's security was given a similar effort.
I am not surprised at all. Technical debt, periodic refactors or security analyses are very low on the list of things that get rewarded, and shiny new projects are high. That's until a security incident takes place, then people are reprimanded and things will be okay for a month or two.
Until that's fixed, things like this will keep being commonplace.
It's still an option in my modern Web browser. A site admin can choose not to track me if they wish - my software is making my wishes clear and in according to a standard.
Worth noting that the spec was abandoned in part because enabling it actually makes you more likely to be tracked. It's another datapoint that can be used to fingerprint your device.
If there were legal basis for that header saying "let me use your website without being tracked" and the website owner having to do so, then it would be worth it to enable it - but, in practice, having it enabled becomes a fingerprint datapoint, and since 99% of traffic doesn't have it enabled, it becomes a pretty reliable one.
I leave it enabled out of the principle, and as a data point. Every person who has such a feature enabled is sending a message: we do not want to be tracked. It's a blip in the dark, but you can't stop the signal.
I'm much happier with the GDPR idea where tracking is not allowed by default and the website has to obtain informed consent first. Of course in practice the implementations still need to be improved - both through enforcement (many of the consent dialogs are not legal), through additional standards and legislation to require them (you should be able to set your browser to automatically deny all requests) and through continuing to shame companies and individuals for their excessive tracking (at least the consent dialogs tell you about the extend now).
I think if you'd ask most people if they want to be tracked they'd say no. Of course you could word things differently or promise benefits to manipulate the answers, but just for "do you want to be tracked" the answer will typicall be NO - and some more progressive privacy laws like the GDPR recognize that by making no tracking the default.
Setting DNT to true by default is certainly more in line with user's wishes than the previous assumption by the tech industry to just grab whatever info they can.
A standard that practically no one cares about and is actually used by others as another flag to identify you since most people don't have it turned on. Ideology is great but on the web some of us choose to be pragmatic.
I doubt Automattic runs Gravatar for free out of the goodness of their hearts. It can track in two easy ways: firstly the user that requests different avatars, as they move through the web and different sites, and secondly the users whose avatars are requested, as to what services they use. Their privacy policy allows for this at a quick glance.
Personally I proxy Gravatar requests through my web service's backend to get rid of the first type of tracking.
According to their own privacy policy the information they collect and share (read sell) includes "online identifiers", internet and network activity, device activity including cookie information, "other device identifiers", IP addresses, and geolocation data
How much do non-giants really need one? Sure, they have their advantages, but are we really at the point where every website absolutely requires their use?
Depends on what you mean by 'absolutely requires'. With search engines now emphasizing mobile experience and load times, CDNs are more necessary than ever if a site wants to appear/rank in search results.
Isn't self hosting actually a huge legal cluster fuck liability?
And as much as I hate Google I think they have better security practices than most websites out there.
A CDN? If you load JavaScript from it, then it has about the same "power" as a server hosting the HTML of a static site. So yes, there's some potential for misuse, but unless you're implying that hosting your own website is a bad idea from a liability perspective, I'd say it's safer to host your own JS instead of using multiple CDNs as many sites do.
The fact that it took Techdirt, A site which supposedly is read by all leading Big Tech CEOs this long to do away with Google Ads shows how helpless nascent independent writers who have a fraction of audience of that of techdirt are when it comes to monetization without Ads.
I'm myself in that category, I know that my blog has enough visitors to break even at least the running cost with Ads but I could not pull myself to do it when I personally hate Ads, Feel its disrespectful to attention and go all the way to ensure I don't see one(Pi-Hole, UBlock etc.).
My content are distributed randomly across several interests and I don't have regular schedule due to health reasons so a patreon wouldn't work. Paypal donations are blocked for my country. So the best bet is 'Buy Me a Coffee' but then I have to pay both Stripe + BMaC commissions, I have to register a business to be able to get International payments on Stripe anyways.
So a direct stripe payment or similar PG is the only option to monetize open content for this scenario and since tax id is involved due to Intl. payments it's not a donation.
I've never minded reasonable ads, but I don't want tracker ads and cookies. I know people who produce good content deserve at least a chance at making a buck, I just don't want to sacrifice my privacy for it though. So serve up ads related to the article content or generic random ads I suppose.
I agree, But advertising agencies which offer advertisements without tracking are negligible because their competition produces more conversion so obviously advertisers flock to them. Services like ethical ads are available only to select large open-source projects as the advertisers target niche audience.
Affiliate links when the content isn't exactly a review of the said product/book (to avoid conflict of interest) do work in this context to some extent. But the current referral system is a mess with each product/service tying up with different middlemen or having their own system, arbitrary approval, payment issues, publishers gaming the system etc.
I have started to track the need-gap for 'Single referral system for all services' on my problem validation platform[1].
I agree, I think this is why tracking/cookies that do so should be illegal. Then at least the big guys like google will stop doing it. You'll never stop all of it though.
The message is that Techdirt is without Google Ads, Tracking Code or analytics.
This may be a set in the right direction but.....
The same page on which they posted this is running CSS from fonts.googleapis.com, as well as embedded frames from soundcloud.com & w.soundcloud.com which I also believe is google owned.
It's not, but the implication is that they're touting their effort to be "much more protective of the privacy of Techdirt's readers" as they begin self-hosting their tracking tools, while reader data is still being sent to SoundCloud, a private third-party with a heavily-vested interest in user tracking.
They're working on it. If you don't read TD regularly, you're probably not aware of how long this has been going on. They're trying to survive independently without selling users out, and that is not easy.
> So, count me a little skeptical
Be as skeptical as you want, they are transparent about what's going on. A decent background piece specific to the Google piece of this madness explains part of it here:
I evaluated both Plausible and Matamo as privacy preserving analytics options for my blog, and ended up rolling my own solution for cost reasons. Cloudflare workers and Pulumi make this pretty trivial to self host.
Maybe like letsencrypt we need letsanalyze that lets you easily self-host your analysis and generates pretty Material Design reports with nice fonts and touch friendly graphs.
We were discussing with them adding https://www.ethicalads.io/ which is our new non-tracking ad network. Sadly we're still only focused on developers, and Techdirt's content is very wide ranging, so we were only able to propose running ads on a subset of their content currently. We'd love to be able to support them when we are able to expand our audience to a larger tech-focused site in the future.
Totally! We're still working on firming up our existing niche, but looking to expand. You can subscribe to our newsletter to keep in the loop (~2x posts a month, sub in the site footer) and see our journey, but I'll also add you to a Trello card for folks to reach out to!
I don't believe carbon does the niche targeting quite as well, but definitely a similar concept. Historically they're allowed third-party JS and images in the ads, which we didn't want. I'm not sure if that's changed though.
We transitioned away from Google ads ~7 years ago.
One solid strategy is to review which companies are advertising on the site via Google Ads and approach them directly. Because of Google’s hefty margin, it can be a win-win for both the publisher and advertiser.
Ironically, the entire Techdirt site is blocked for me by one of my content blockers. I don't actually know what Techdirt is or whether it's worth unblocking.
Techdirt is a tech news site with articles that are longer than soundbites. They cover a range of tech topics with a focus less on Company X just released this great product Y.
I’d look into who sets the rules on your content blocker. I can’t think of any reason why Techdirt might be blocked beyond the common tracking that we are discussing here and which they are trying to get rid of.
Thanks (to you and others). A post elsewhere in the comments pointed out that removing `ads` from the url works - which is the same for me. For the record, I'm using AdGuard in Safari, with DNS from AdGuard Home on a Raspberry Pi with upstream DNS set to NextDNS. I'm very happy with this setup, can't remember the last time an ad or tracker slipped through.
Mike Masnick/Techdirt's political lobbying arm, Copia Institute, still proudly promotes their sponsorship from Google on their front page: https://copia.is/
Now, I don't know what kind of scripts are being loaded from these sources but things are looking refreshingly spartan, considering techdirt is an online publication
For comparison, arstechnica usually has like 18 sources, 10 of which are trackers
Tech sites have been in a bad state for over a decade at least, after the arms race that occurred that 'freshened up' these websites from what they were around 2005 or so. Thankfully they still have RSS feeds or I can use services to scrape an RSS feed from them.
The best part is when they even tell you what is going to be served even if you do pay for premium. No ads but most tracking remains. It's mostly about image
Ah, I didn't realize that the modifier "Google" also applied to the phrase "tracking code" in the post title. Also, my understanding is that the GDPR notification is only required for tracking cookies, not all cookies.
For example, I buy hot rod magazines to get the advertisements selling hot rod parts in those magazines. I don't buy them to read toothpaste ads. I even pay money to get those ads.
But if I was visiting a site about dental hygiene, I'd be interest in the toothpaste ads.
I've tried to configure context-sensitive ads from Amazon on my web site to be books on programming. But Amazon would insist on showing:
1. batman movie ads
2. the same ad for a C++ course over and over and over and over
So I wound up removing those from the site. Instead, I created a list of good programming books myself, and would randomly cycle through that list to show an ad.
It would be so much better if I could enable:
"Only show books from this category."
as my aim is to only show advertisements that would be of interest to the people reading the site.
Google has context sensitive ads, and one would think this would work. But it doesn't. For example, I have a page on the American Revolution. Google's context sensitive ads are usually for travel agencies, presumably based on the place names on the page. Actually relevant ads would be for history books. I tried meta keywords, etc., nothing worked. I gave up. Google's context algorithms are worthless and the site owner should be able to provide guidance.
The problem is that most people spend most of their time on "Which Harry Potter character am I" and stuff like that and they're already Harry-Potter-saturated at that time. Contextual ads have their place, certainly, but chasing ROI ultimately does lead you to target segments based on other things.
For instance, if you were browsing high-end networking hardware and you decided to go browse chicken recipes, it's still better to show you network device ads simply because even if you have a 1/1000th chance of converting on them, you'll be worth more than buying chicken.
Note that I'm willing to pay to browse relevant ads, as are a large number of people who buy specialty magazines.
But there's no advertisement broker that provides such a service to web sites. They all follow the google/amazon model of serving irrelevant and annoying batman and travel agency ads, which motivate people to install ad blockers.
Note that (before the intertoobs) there were other products where people were willing to buy ads to browse:
1. the yellow pages
2. newspaper classified ads
3. JC Whitney catalogs
4. Hemmings Motor News
5. all the computer/electronics/hobby magazines
Another method is exemplified by the "Detroit Muscle" car show. It is ostensibly about how to soup up your car. But it's actually an ad platform, as they use various tools and supplies to do the work, and show how to use them. That's what pays for the show and it's why I watch the show. I like ones about how to bend metal and here's the tool that does it. There are shows on how to weld and here's the welder for the job.
Those paper based ads works exceedingly well particularly if you don't know the area. For example, growing up my grandparents knew next to nothing at all about what toys kids were interested in. When they wanted to buy us toys for christmas or a birthday or something, they'd seek out relevant advertising in the form of the Sears catalog toy section, and browse through all popular toys on offer and their prices all in one place. We'd even circle the toys in particular we wanted from that catalog. I can't imagine what they'd do today if they were still around trying to buy things for grandchildren. You can't exactly get a catalog from Sears anymore. You have to go on a wild goose chase on the internet. Its a chore even for me having grown up with the internet.
What the Sears catalog provided, in addition to a way to choose items, was a curated list of items that were chosen to be good (or good enough) in they category. Amazon is more of an unregulated flea market.
If it was about chance of conversion and being worth a lot, why aren't I getting ads for learjets all the time instead of for paper towels? If the learjet ad converted 1/1000000 of times, I'd be worth probably an order of magnitude more buying a $50m jet than buying a $6 pack of scott towels even with the low conversion rate, but the scott towel ads are what follows me around the web after searching once for them, and these are ads I continue to ignore because what influences me with my paper towel choices isn't an idylic image of some smiling person wiping a table, but how much they cost at my grocery store.
It just seems like conversion rates would improve no matter what you are selling if you just spend a little bit more time thinking about advertizing in relevant areas vs the firehose approach that is done now, banking on the large numbers of it all to overcome the small conversion rate. If Scott wants to sell more paper towels to me, they could almost guarentee my business if they stopped advertising via irrelevant ads, and advertised in relevant spaces such as if they included a coupon for their paper towels with my weekly grocery store mailer that made their offering cheaper than their competitors.
All of the research I have seen (of which there is very little) shows that targeted advertising offers only a marginal benefit (low single digit percentages). Is it worth alienating your users and being generally creepy for such marginal gains?
This post reminds me we need to write up a full blog post on how it works, and how we're seeing great results from it. We're trying to prove the model works, so always happy to have folks support us!
This looks like a good initiative. May I suggest also providing a mechanism for the site owner to specify, on a per-page basis, the topic/category/etc for the ads. The site owner will know what their viewers are interested in more than any automated system.
Also, some "negative" requirements, like "no batman, no toys, no games, no movies".
This would assume a very wide amount of supply that could possibly fill the demand for said predefined niches. Plus all of the magic around cost, conversions, and other metrics that ad companies use to select ads for a site.
As you noted even Amazon was showing the same C++ ad. This might be because it was all they had that was relevant (plus high spend out-competing others) and there wasn't enough other publishers paying to advertise books/products in other related categories.
It wasn't the C++ ad that was the problem in this case, as it was programming related. The problem was that same ad was run for every page every day. It wasn't long before I grew to hate it.
The affiliate money from them is not enough to interest me. What does interest me is providing ads that my readers would like to see.
Maybe I'm a unique snowflake, but relentlessly showing me the same ad makes me hate it. I don't listen to internet radio because each station runs the same 3 ads all day. I just can't stand it.
The world's greatest invention was Tivo, so I no longer had to suffer decades of Cal Worthington and his goddamned Dog Spot any more.
The second was car radios with a USB plug so I could plug in a stick of my music instead of ads-for-that-wretched-jewelry-store that I shall not name nor ever buy from. 40 years of those ads. barf.
May I suggest a "categories" option. Ebay and Amazon have categories one can restrict searches to, for example. Just that will go a long way.
Keywords tend to have multiple meanings, and advertisers won't really know what keywords to use to represent a particular category. For example, I'd like to specify
"books and programming"
so only ads that are for programming books would appear.
Based on the wording from your website it looks like it will suffer the exact problem GP is talking about. Can Ethical Ads reliably display ads for sim racing / gaming gear on a page about F1 2021 the game, and motorsport merch for a page about Formula 1 the sport?
Don't rely on your algorithms. Let the page owner ultimately decide what gets advertised.
My hand-curated list of well-regarded programming books that my programming site cycles through has been a success. Not a single person has ever complained about those ads.
I view a proper set of ads that are relevant to the site content as a service to the user, not an annoyance.
I (and presumably my viewers) would appreciate a way of specifying this to your ad servers. I would also appreciate a way of restricting ads to being text/picture/link and nothing more.
I agree with your argument, but I don't think your example is good because I mostly read magazines while waiting for a doctor OR dentist. So if your example is cherry-picked from real life, that ad was there for a reason.
My uBlock Origin blocks 3 requests, AdBlockPlus blocks 5 and Privacy badger blocks another 3. en.wikipedia.org and news.ycombinator.com have 0 on all three filters.
Is anyone here familiar with Matomo? I noticed the linked page attempts to run it before seeing it mentioned, because Ghostery blocked it.
It's a “Free software alternative to Google Analytics” “which can easily be configured to respect your visitors' privacy”[0] (emphasis mine). But can it be configured in a way so you can claim “without any tracking code”? I wonder whether the claim in the headline was supposed to be merely “no third-party tracking”. [Edited to add:] Waiiit a moment, is the intended meaning simply “no Google tracking code”? (I don't think the article addresses this anywhere; hope I didn't just miss something.)
Or maybe I'm supposed to know that “no tracking” is jargon for something like ”attempts to not store PII” in this industry? I doubt that it can be configured such that, for example, unique visitors are statistically estimated instead of counting unique IDs from tracking cookies — but I'm too lazy to research that in detail, so it would be interesting to hear from someone who knows more.
Well, the only thing you can do is attempt not to store PII and only pull very limited information from the browser, ideally just the utm tags, referrer, and maybe some available api feature flags (eg. say you want to roll out webauthn and want to figure out how many of your users' browsers have that capability). Unless you trust that the site has `access_logs off`, tracking can just be done by analyzing those access logs even for users with noscript.
Google Ads is so deep into tracking that it defeats its own purpose.
I search for Product A on Google or Amazon and am immediately inundated with ads for the same product on every page I go that has Google ads
What is worse is that the ads continue even after I purchase the product / service. So I get ads for a mobile phone I searched for and bought in Amazon (and all its accessories) for months after I have already purchased it.
Can someone tell me why Techdirt is not featured often in the Google front page results when compared to other media? Is it because they originally had summary to 3rd party content and so Google ranked them low for not having original content?
118 comments
[ 4.4 ms ] story [ 140 ms ] threadhttps://www.techdirt.com/articles/20210726/09441047251/techd...
Same here, with Adguard.
The problem is with the title part of the url, since the title contains the word "ads".
I did an experiment: If you remove the words after "-google" in the url, it still works. You could probably remove the entire title of the url and it will still work by article id... So here's a link that will work with Adguard enabled:
https://www.techdirt.com/articles/20210726/09441047251/techd...
Regardless of ownership, you’ll run into same problems almost everywhere if you publish articles critical of your advertisers. That’s nothing to do with Google and just the market working as intended/expected.
the correct complaint, IMHO, is that the ad networks themselves, regardless of who powers it... are shady. singling out google is not entirely fair.
Or just for kicks?
To my knowledge, Google's basic nature has been known for quite a while, and people who care have lined up on different sides depending on their ideologies and self-interests. What has changed in the last 6 months that makes so many companies change their opinions about working with them? Is it just that public opinion has crossed some tipping point where it makes sense to gamble on a temporary revenue hit in order to be on the "right side"? That seems unlikely, just because I don't know a lot of people who weren't skeptical of Google a year ago who are now.
Is it related to FLoC? That seems like a wonky issue that's under most people's radar. Is it the EU antitrust hearings? Also seems unlikely.
Honest question: why now?
I would have guessed it would be most of their readers blocking easy-to-block ads, and the low payout, but it seems capricious policy enforcement is happening across Google and Youtube especially.
More cynical dynamic is virtue signaling, which can pretend to be the former.
You know, the sort of thing that could potentially kick off a frenzy… as unlikely as that seems.
The way advertising companies used to treat online advertising a decade ago was that it didn't matter where your ads were placed, as long as they were going in front of people who were interested in them. That's why you had large ad exchanges with third-party tracking tools to build up behavioral profiles, because it was the best approach for linking advertisers and publishers. Publishers generally didn't care about who was advertising on their pages and advertisers generally didn't care about what pages their ads wound up on. This is, of course, a generality; there's always been picky advertisers and/or publishers, but the core point remains: the business worked because inventory was fungible.
And then advertisers realized that their ads were showing up on videos of Klu Klux Klan rallies and other far-right nonsense.
Well, they didn't so much realize, as much as they were publicly embarrassed when journalists started writing news stories about it. Consequently, you saw multiple waves of brands making a show of pulling all their advertising until exchanges implemented better controls for brand association and advertiser safety. Of course, this means that exchanges wound up over-enforcing ad safety, constantly harassing publishers that might be on the line. Journalistic outfits were unusually impacted, because their content is rarely advertiser friendly, even though they're not the right-wing nutters that started this mess.
This problem isn't related to trying to keep right-wingers out of ad money, though. There's a sort of converse Adpocalypse that happened to publishers, called malvertising. This can take many forms - from fake popups for scam tech support companies, to just outright dropping a zero-day exploit on your victim. This causes reputational problems for publishers, since safe-browing services will flag your site (basically making it unusable without lots of scary warnings, and heavily punishing it in search engines) if it starts dropping drive-by downloads on users.
So, advertisers are unhappy because of bad publishers, and publishers are unhappy because of bad advertisers. What about the users? Well, they're installing ad blockers in record numbers because they don't want to get hacked by malicious ads. Also, many of them aren't actual users; the online advertising business has had click fraud problems way before malvertising or brands getting cold feet over what content they're paired with. That's actually part and parcel of why online advertising is so tracking-heavy; it lets them detect the difference between an actual reader and someone just clicking on their own ads all day to defraud advertisers. Guess what? Users don't like ad tracking all that much either (especially when exchanges are so directly in-your-face about it.)
In any fungible market, the presence of "bad" commodities drives out "good" ones. What you have now is a toxic market, one that Google more or less owns outright because everyone else has been driven out of business or acquired. Google isn't the cause of the toxicity, but they benefit from it regardless, since Search display advertising will remain brand- and user-safe in ways publisher advertising isn't. In this new market of TV-like, non-fungible advertising; Google is no longer a business partner. They are a hostile, dogged competitor paying them ever-shrinking table scraps.
• https://www.stefanjudis.com/notes/say-goodbye-to-resource-ca...
• https://wicki.io/posts/2020-11-goodbye-google-fonts/
Their "self-hosted" Plausible content is pulled from tics.techdirect.com, which resolves to 46.101.161.209 and shares a reverse DNS with custom.plausible.io. Can't tell who runs the web server on the other end of this, so it may well be self-hosted.
The site also dropped a first-party one-year cookie, and let a third party drop a week-long cookie.
My browser is set to send Do Not Track requests, so my explicit wishes are still being flagrantly ignored.
There is still a lot of third-party tracking going on here, as well as execution of remote code under the control of third parties. It's remains a privacy and security clusterfuck but it's better than it was.
Wait, they're still serving things via rawgit? They started shutting down in 2018! That service has been deprecated for three years. That sounds like an attack that is just waiting to happen.
> GitHub repositories that served content through RawGit within the last month will continue to be served until at least October of 2019. URLs for other repositories are no longer being served.
> If you're currently using RawGit, please stop using it as soon as you can.
Until that's fixed, things like this will keep being commonplace.
That looks like it's Plausible's "they host it" system.
That behaviour is certainly not software making their wishes clear in the matter.
Setting DNT to true by default is certainly more in line with user's wishes than the previous assumption by the tech industry to just grab whatever info they can.
> are now using both Plausible and Matomo (self-hosted),
Is it possible they only meant self hosted Matomo, not Plausible?
I have a dollar that says most websites you visit do not respect this setting unless they don’t use trackers at all.
That's not tracking right?
Personally I proxy Gravatar requests through my web service's backend to get rid of the first type of tracking.
I'm myself in that category, I know that my blog has enough visitors to break even at least the running cost with Ads but I could not pull myself to do it when I personally hate Ads, Feel its disrespectful to attention and go all the way to ensure I don't see one(Pi-Hole, UBlock etc.).
My content are distributed randomly across several interests and I don't have regular schedule due to health reasons so a patreon wouldn't work. Paypal donations are blocked for my country. So the best bet is 'Buy Me a Coffee' but then I have to pay both Stripe + BMaC commissions, I have to register a business to be able to get International payments on Stripe anyways.
So a direct stripe payment or similar PG is the only option to monetize open content for this scenario and since tax id is involved due to Intl. payments it's not a donation.
Affiliate links when the content isn't exactly a review of the said product/book (to avoid conflict of interest) do work in this context to some extent. But the current referral system is a mess with each product/service tying up with different middlemen or having their own system, arbitrary approval, payment issues, publishers gaming the system etc.
I have started to track the need-gap for 'Single referral system for all services' on my problem validation platform[1].
[1] https://needgap.com/problems/274-single-referral-system-for-...
This may be a set in the right direction but.....
The same page on which they posted this is running CSS from fonts.googleapis.com, as well as embedded frames from soundcloud.com & w.soundcloud.com which I also believe is google owned.
So, count me a little skeptical
https://duckduckgo.com/?q=google+owns+soundcloud&t=brave&ia=...
This is erroneous, as soundcloud is not google owned.
> So, count me a little skeptical
Be as skeptical as you want, they are transparent about what's going on. A decent background piece specific to the Google piece of this madness explains part of it here:
https://www.techdirt.com/articles/20200810/11335745081/why-a...
(I've met with Mike a couple times, but otherwise have no connection to them, other than being a long-time reader.)
Also, Google does not own Soundcloud.
Here's the Show HN that never picked up steam: https://news.ycombinator.com/item?id=27175347
Sign up a few major brands that market to everybody, such as Ford and Pepsi. You tend to not see those on Google. Offer 4-second GIFs.
Get acquired by IAC.
2. Go to Ford.com and build a Mach-E configuration
3. wait a few days
4. load some Google Ads
Ford is definitely an advertisers on Google Ads. I get them every time I go back to check out the Mach-E or other upcoming EVs from them.
One solid strategy is to review which companies are advertising on the site via Google Ads and approach them directly. Because of Google’s hefty margin, it can be a win-win for both the publisher and advertiser.
I’d look into who sets the rules on your content blocker. I can’t think of any reason why Techdirt might be blocked beyond the common tracking that we are discussing here and which they are trying to get rid of.
1. techdirt.com
2. cloudflare.com
3. fontawesome.com
4. s3.amazonaws.com
5. soundcloud.com
Now, I don't know what kind of scripts are being loaded from these sources but things are looking refreshingly spartan, considering techdirt is an online publication
For comparison, arstechnica usually has like 18 sources, 10 of which are trackers
For example, I buy hot rod magazines to get the advertisements selling hot rod parts in those magazines. I don't buy them to read toothpaste ads. I even pay money to get those ads.
But if I was visiting a site about dental hygiene, I'd be interest in the toothpaste ads.
I've tried to configure context-sensitive ads from Amazon on my web site to be books on programming. But Amazon would insist on showing:
1. batman movie ads
2. the same ad for a C++ course over and over and over and over
So I wound up removing those from the site. Instead, I created a list of good programming books myself, and would randomly cycle through that list to show an ad.
It would be so much better if I could enable:
"Only show books from this category."
as my aim is to only show advertisements that would be of interest to the people reading the site.
Google has context sensitive ads, and one would think this would work. But it doesn't. For example, I have a page on the American Revolution. Google's context sensitive ads are usually for travel agencies, presumably based on the place names on the page. Actually relevant ads would be for history books. I tried meta keywords, etc., nothing worked. I gave up. Google's context algorithms are worthless and the site owner should be able to provide guidance.
For instance, if you were browsing high-end networking hardware and you decided to go browse chicken recipes, it's still better to show you network device ads simply because even if you have a 1/1000th chance of converting on them, you'll be worth more than buying chicken.
But there's no advertisement broker that provides such a service to web sites. They all follow the google/amazon model of serving irrelevant and annoying batman and travel agency ads, which motivate people to install ad blockers.
Note that (before the intertoobs) there were other products where people were willing to buy ads to browse:
1. the yellow pages
2. newspaper classified ads
3. JC Whitney catalogs
4. Hemmings Motor News
5. all the computer/electronics/hobby magazines
Another method is exemplified by the "Detroit Muscle" car show. It is ostensibly about how to soup up your car. But it's actually an ad platform, as they use various tools and supplies to do the work, and show how to use them. That's what pays for the show and it's why I watch the show. I like ones about how to bend metal and here's the tool that does it. There are shows on how to weld and here's the welder for the job.
It just seems like conversion rates would improve no matter what you are selling if you just spend a little bit more time thinking about advertizing in relevant areas vs the firehose approach that is done now, banking on the large numbers of it all to overcome the small conversion rate. If Scott wants to sell more paper towels to me, they could almost guarentee my business if they stopped advertising via irrelevant ads, and advertised in relevant spaces such as if they included a coupon for their paper towels with my weekly grocery store mailer that made their offering cheaper than their competitors.
https://techcrunch.com/2019/05/31/targeted-ads-offer-little-...
This post reminds me we need to write up a full blog post on how it works, and how we're seeing great results from it. We're trying to prove the model works, so always happy to have folks support us!
Also, some "negative" requirements, like "no batman, no toys, no games, no movies".
As you noted even Amazon was showing the same C++ ad. This might be because it was all they had that was relevant (plus high spend out-competing others) and there wasn't enough other publishers paying to advertise books/products in other related categories.
Just something to consider.
The affiliate money from them is not enough to interest me. What does interest me is providing ads that my readers would like to see.
Maybe I'm a unique snowflake, but relentlessly showing me the same ad makes me hate it. I don't listen to internet radio because each station runs the same 3 ads all day. I just can't stand it.
The second was car radios with a USB plug so I could plug in a stick of my music instead of ads-for-that-wretched-jewelry-store that I shall not name nor ever buy from. 40 years of those ads. barf.
Keywords tend to have multiple meanings, and advertisers won't really know what keywords to use to represent a particular category. For example, I'd like to specify
so only ads that are for programming books would appear.Don't rely on your algorithms. Let the page owner ultimately decide what gets advertised.
Good luck to you and your endeavours ;-)
My hand-curated list of well-regarded programming books that my programming site cycles through has been a success. Not a single person has ever complained about those ads.
I view a proper set of ads that are relevant to the site content as a service to the user, not an annoyance.
I (and presumably my viewers) would appreciate a way of specifying this to your ad servers. I would also appreciate a way of restricting ads to being text/picture/link and nothing more.
It's a “Free software alternative to Google Analytics” “which can easily be configured to respect your visitors' privacy”[0] (emphasis mine). But can it be configured in a way so you can claim “without any tracking code”? I wonder whether the claim in the headline was supposed to be merely “no third-party tracking”. [Edited to add:] Waiiit a moment, is the intended meaning simply “no Google tracking code”? (I don't think the article addresses this anywhere; hope I didn't just miss something.)
Or maybe I'm supposed to know that “no tracking” is jargon for something like ”attempts to not store PII” in this industry? I doubt that it can be configured such that, for example, unique visitors are statistically estimated instead of counting unique IDs from tracking cookies — but I'm too lazy to research that in detail, so it would be interesting to hear from someone who knows more.
[0] https://github.com/matomo-org/matomo
I search for Product A on Google or Amazon and am immediately inundated with ads for the same product on every page I go that has Google ads
What is worse is that the ads continue even after I purchase the product / service. So I get ads for a mobile phone I searched for and bought in Amazon (and all its accessories) for months after I have already purchased it.
My mind automatically ignores every ad.
But now they seem to have only original content.