Hacker News new | past | comments | ask | show | jobs | submit login
Google Adsense is transitioning to per-impression payments for publishers (blog.google)
142 points by maoro on Nov 2, 2023 | hide | past | favorite | 116 comments



Google's premium ad network for professional publishers is AdX which is run via the google ad manager product and has been on a CPM basis for over a decade. Only small sites are on adsense which pays CPC.

The net change here is probably almost nothing, just the smaller sites that never use google ad manager will see the change but any publisher of note will have been operating with this for as long as they can remember.

With click through rates continuing to decrease it's likely they needed to make this change to keep the long tail sites happy and generating some revenue, they would back out the CPC to an effective CPM anyway.

Me:13+ years in digital publishing and advertising.


Cost Per Click and Cost Per Mille (thousand impressions) for anyone else who was wondering.


Me thinks that this is related to the youtube change. Before only ads that were clicked on counted in earning revenue for Google and YouTube. With this change every inpression is worth that much more.


> Before only ads that were clicked on counted in earning revenue for Google

This certainly wasn’t true 10 years ago. I’d have low traffic pages go without a click but still earn some cents in revenue that day running Adsense (effective CPM would be a rough baseline of ~ 1/10th what a page with clicks that day would earn).


As a general market trend, why are click through rates decreasing? Is it because of the availability of information without going to a site?


> As a general market trend, why are click through rates decreasing?

Because the advertising industry has structurally educated users to try their hardest to ignore the ads, resulting in every more obnoxious ads, which are then even more forcefully ignored. And that in turn gave rise to adblockers, which are now more or less a requirement if you don't want to lose your sanity while browsing the web.


> which are now more or less a requirement if you don't want to lose your sanity while browsing the web

For a long time I wondered why people said this. I don't use ad blockers and didn't feel it was that bad.

Then 2 things made me understand. First, I pay for YouTube. If you don't and don't block their ads, they seriously test your patience as you browse. I tried it for less than an hour before I couldn't take it anymore.

The second was looking for torrents and hacks (it was for a legitimate and unambiguously legal purpose too, no gray area, but long story!). Those sites are literally impossible to use without ad blockers. Same thing for tools related to diagnosing PC issues. It's all ads over ads over scams and trying to get you to install some adware as you navigate the site trying to install the actually legitimate tools.

I very rarely do either of those things (YouTube without subscription, and navigating the "gray" web), so I never realized just how fucking awful it can get.


Hell, I'm a YouTube Premium subscriber and I still have an ad-skipper for all of the in-video promotions. It's not quite as bad as direct YouTube ads have gotten but it's still a noticeable change in itself.


I dont think you even need to consider gray web stuff. Your average news site is awful and unisable without an ad blocker.


There seem to be some logical considerations that make the growth of advertising self destructive for advertisers. The big one is that the more ads one sees, the less likely they become to interact with any given ad. There will also likely be an increased ability of users to ignore the ads they do see. As advertisers see returns from advertising decrease, the costs for advertising will gradually trend downwards. This will enable even more advertisers to start advertising.

It's almost like a sort of reverse network effect. The bigger you get, the less desirable your product becomes. And the advertising industry's getting quite large.


Also there are more "branding" ads which are placed just to get the company's logo in your face and don't necessarily need a click to do their job.

If I want to advertise my new Windex(tm) flavored Mountain Dew[1] then all I need is people to see it. They're not likely to buy it from a click, but I want them to remember it next time they are at the grocery store, and I'll pay to shove it in their browser.

[1] Not currently available in retail stores


It could be because of number of ads shown per query. As number of competing ads goes up CTR goes down.


I'd also like to see a source for that claim. Been working in digital advertising for 10 years, so I'd be seriously interested. We've changed so much in our advertising over the years I wouldn't be able to tell whether that trend really exists.


I personally just open a new tab and search for the same (i wonder if google can connect those 2 dots to count it as conversion)


> For years, AdSense has been transparent about the fee we charge for our service, which is consistent with industry rates. When publishers have chosen to use AdSense to monetize their content, they have kept 68% of the revenue.

Unless, of course, when you go to cash out you run into their absurd KYC that will arbitrarily get rejected (you only get 1 chance), or they cancel your account for reasons, and there is absolutely no recourse.

They are happy taking your business before that though, of course.


Happened to me too when i was a student, earned $300, tried withdrawing and they just froze my account, for "fraud clicks". It's Google's long tail business model to not filter fraud out on the go and instead just lead small site owner on. On a global scale I expect billions in additional revenue, but no global court to challenge Google with a class-action. Maybe someday...


They got me on a hobby site for about $5k. No real reason given, just rejected my ID verification with no appeal possible. hundreds of similar stories out there


I'd go to small claims over $5k.


There was an era when Google’s slogan was ”Don’t do evil.”

It was so long time ago that most of have not even heard about it today.


Don't be evil.

Funny how even you don't really remember it

https://en.wikipedia.org/wiki/Don%27t_be_evil


Don't be [caught doing] evil.

Someone just paraphrased that in the company memo.


Yeah it's not been a thing a google for at least a decade. When I went to work for them the onboarding did not have that phrase anywhere in any of the documents (onboarding, training, or orientation) - and I explicitly searched for it.


They didn't ask you to read the Code of Conduct? It's been in there from the beginning.

https://abc.xyz/investor/google-code-of-conduct/


They were noticed publicly removing it in 2018 (which was reported on), so I'm guessing I came on during a period it was gone, and they have since re-added it.


Most of the public reporting was incorrect. The phrase never left the Google code of conduct, but it did move to a less prominent position, and the closest thing that has ever been in the code of conduct of the parent company Alphabet since its creation in 2015 is “Do the right thing”.

“Less prominent position” does not mean “zero prominence”. A mention in the final sentence is not anywhere near as prominent as it was before the change, but that’s still more prominent than something like hiding it in the middle.

Now, their adherence to the phrase has certainly decreased over time and was never perfect. That’s a separate matter.

Disclosure: I used to work for Google, but not since before Alphabet was created (which was some years before this Google code of conduct change). I never had any involvement in the decisions over this motto or the changes to where it shows up, beyond of course trying my best to adhere to it in my own work.


It's listed in both January 2018[1] and December 2018[2]. They moved it from the opening line to the closing line, but they did not remove it.

https://web.archive.org/web/20180117100018/https://abc.xyz/i...

https://web.archive.org/web/20181225060308/https://abc.xyz/i...


Nonsense. It’s been in the employee handbook (which you had to sign) and in the “10 things we know are true” for like 20 years


Except it isn't true and likely never was. Good PR though!


That is unfair. The engineer who suggested the phrase and the others around at the time were being honest.

But organizations change as they grow.


Engineers don't set policy.


At the time they (Larry and Sergey) did.


Yes, and look what happened. It's not as if they have disappeared, they handed the reins to a guy who walked all over that and got away with it. If they really cared they would have put a stop to that and the uncountable other privacy and tracking issues that Google has been up to over the years.


I do not recall seeing it, and I do recall looking for it explicitly. Which means even if it were there - which given this was the era in which they were removing stuff like this in order to win pentagon contracts seems completely plausible - it was not considered important enough to highlight or to put at the forefront of employee information.


I earned a couple hundred with them too when I was young. Never claimed it. They eventually released it to my states unclaimed property system and a decade later I got it from that.


> It's Google's long tail business model to not filter fraud out on the go

Well the honest answer is that it's harder for clickfraud farms to hyperoptimize against detection algorithms on-the-fly this way. This makes it harder for them to figure out exactly what pattern is flagging the algorithm.


This happened to me when I was in college and cost me a few hundred dollars. It caused so much stress, and no human was willing or able to help me. As far as I'm concerned Google robbed me.


Small claims court? May be too late now of course.


They have several thousand of my dollars. Every several years I get a notice from an auditor that I have money and I just need to login and get it. Except I can't login, and I can't get it.


All of these anecdotes smell like a class action lawsuit. Perhaps also criminal charges? It's theft. Felony grand theft, in your case. Maybe their pretext is enough to muddy establishing intent? IANAL, it just sounds absurd that Google can skate by while doing this.


It happened to me for no good reasons in 2008.

Nowadays I have a side projects with Adsense on it, using another Google account, which I have to use in incognito mode because otherwise it links me to the other account that I'm still using and rejects me.

Also, for a service that processes billions, it still does not offer a developer mode so when your implementing the ad script:

- You're not always getting ads so can't be really be sure your code and UI are OK.

- You risk being penalized / kicked out because you click on an ad n times by mistake

Edit: also there's a semi-scam that has been running on Adsense for years. Those "3 steps to get your video/software/etc click here" ads. There is a lot of variant.

IIRC they trick people into subscribing to a fake service through your phone bill.

They're very hard to block as a publisher as they come from A LOT of accounts.


Yeah, the business model is remarkably similar to the credit card fraud reshippers. Keep your mules busy until is time to pay out and then ghost them.


Wasn't CPM the norm before the market went to CPC because it was easily gamed by publishers?

Also it's funny for them to say "Publishers gets 68%" which sounds huge but actually translates to a staggering 32% fee.


I think YouTube is 50% !

But Adsense was capable of earning so much more than what most could do themselves that I was happy to “pay” them when I ran a semi-popular site.

Though CPM can be gamed as well: put the ad at the bottom instead of the top. At the right instead of the left.

Can be gamed by publishers too: create an ad that users see but don’t click on, but act on anyway. E.g. an ad for a cola and then you drink one from your fridge.

I thought Google basically blended it all together when deciding which ad to present: a poorly clicked ad will need to pay more per click to get displayed.


I was working at Yahoo Travel from 2004-2011, and touched a lot of ad stuff, although some of it was indirect.

Media/Display ads were normally CPM. Text ads were nearly all CPC (regardless of adwords, yahoo's search marketting, or y!travel's custom advertising market). Some display ads were CPC. There was a small amount of ads where it was CPA (cost per acquisition, advertiser only paid if the user purchased). CPA seemed like it was growing, but details of sales tracking and trust between parties probably gets in the way. Ocassionally we'd have some deal where we would have to put in links that might look like ads, but were unpaid.


CPA seems like the fairest model from my laypersons vantage point.

Is there a reason this isn't more popular on its face? Only paying for an ad when it works seems like you could charge more for the when it works part as a service (in this case Yahoo Ads or Google Ads) but it allows businesses to run ads at a higher rate if they're willing to give up more money on impression (where impression = someone buys)


The big issue with CPA is attribution. It's one thing if you see the ad, click through, and make a booking in a single session. That's easy to attribute, if all the parties involved trust each other to do the tracking. (Parties include the advertiser, the publisher, the ad network(s), maybe even the advertising agency)

But what if you clicked through the ad, and then close the window, but later come back and book through the advertised site? Now we also need to have come to an agreement about 'lookback window'. Maybe a fair lookback window is different for different markets.

But it gets worse. What if you clicked through the ad on your phone, and then booked on your desktop? No tracking. Or even worse, what if it's a brand ad, and you come into the physical store and buy? Definitely no tracking. (well, some people try to get tracking for that, but it's pretty iffy)


Also, tracking is getting increasingly harder.


It shifts risk to the publisher. The publisher has already provided the service of making their user experience in aid of the advertisers goals. Should they get nothing for that? What if the advertiser just made an uninteresting ad? What if the advertiser's products are overpriced?


It requires a lot of tracking and there's a lot of games that can be played as to who actually "made the sale"


AdWords launched as CPM.

Source: I was the lead engineer on the first release of AdWords.


No it wasn't but with a CPM you can actually optimize the ads for better results since the click is the only real measure of successful ad placement. Everything else happens after a click, so getting to the click is the most important target.

As for %. They're bringing the customer, providing the tool, the payments and the collection of the money. All the publisher is doing is placing some code on their site. So 32% is actually a lot cheaper than a publisher hiring their own sales staff, buying their own ad servers, collecting and sorting the payment.

Just so you know.


A lot of ads aren't about a click and direct transaction. FMCG brands run ads just to increase brand exposure all the time, there is no e-commerce environment to buy a can of beer or a tube of toothpaste after the click.


I wonder if it would be cheaper if there were more competitors to google in the market than there are currently.

There are many things that I could outsource that would be cheaper than doing myself (build/buy a car. Buy a cheeseburger as opposed to producing all the ingredients myself.) I am equally interested in what that 32% fee would look like in a more open market and then say if it is cheap or expensive.


There are and have always been lots and lots of options. It's up to the publisher to decide who they work with. But with the scale that Google holds, it's hard to compete with their money as they own the market and have a black box monopoly - there is no way to know what the actual floor of the bid is for instance, its set by Google, arbitrarily.

In other words: They say it costs $5 per click or $25 CPM, they just make that figure up. There is no real market driving that cost. It's 100% made up to maximize Google's profits.


Isn't that exactly what this is addressing?

This is splitting up AdSense such that any third-party buy-side operation that wants to use AdSense as the network is on equal footing with Google's own buy-side using AdSense. If anyone else can run that side of the equation better or for cheaper, they can now do so.


Didn't Barry Dillar go see Sergey and Larry early in Google's beginnings? I think the direct quote was "you are f'ing up the magic" referring to the fact that advertisers don't know what's working, and Google was ruining that. Although it's taken them many years they seem to now be taking his advice to heart.

I looked to cite this but couldn't find it on Google, so I may be mistaken...


https://www.theguardian.com/technology/2010/mar/04/google-ke... looks like 2 different things (barry dillar met them but larry was on phone which he thought was rude, clicks vs impressions was something google was looking for in CEO (eric schmidt passed that test or something))


Google was never actually changing that.

We still largely don't know what half of advertising works and what half doesn't. There's just more metrics obfuscating that reality.

Source: worked on market research focused on identifying what advertising worked and what advertising didn't which Google white labeled


What's the saying? "Half the money I spend on advertising is wasted; the trouble is I don't know which half."

And this quote long predates the internet...


was this solved with conversion tracking long ago?..


Not really. For example, did they see a TV ad which primed them before they saw the banner ad they clicked that led to the conversion?

I recall a conversation with a CMO who was discussing how all their metrics indicated TV ads were a waste of money, so in a few of their DMAs they pulled TV ads and their sales in those areas dropped by double digits.

The additional metric fidelity is a false sense of security in the data. There's a lot more going on to drive someone to purchase than simply the ad tagging that gets picked up. It can be useful in optimizing spend a bit, but should be considered more like correlation than causation.

So you can add budget to things that are converting and see if that linearly leads to more conversions, just don't assume that it's as clear cut as a dashboard makes it out to be - it generally won't be and after a few low hanging fruit gains you'll hit diminishing returns because it's not actually as simple as "person sees banner ad A and clicks = 4% chance of sale."


old cable TV ads, yes, can't be tracked. Ads on HBO Max and Hulu totally can.


Yes. But it goes well beyond just that.

For example, you might have 4% of people who clicked through your banner ad end up purchasing. But 80% of those 4% had previously read a CNET review for your product before seeing that ad. How are you getting that data point? And even if you do, it doesn't necessarily mean putting a link to that review on the landing page for the ad will result in much lift, as there may have been a selection bias where most people swayed by a CNET review were the ones who had already read reviews.

One of the most influential touch points is WoM. So certainly offline product recommendations aren't ending up in your dashboard, and exposure to social media positive sentiment isn't either.

Easily half to 2/3rd of the things someone sees that influence their purchasing decision aren't being picked up by an attribution model for conversion tracking.

And "old cable TV ads" . Still an audience of 72 million for those in 2023.

And cross platform attribution really isn't as straightforward as you seem to think.


Yes, we have different confidence slices: if person clicked on banner, and made purchase in YY mins, it is XX% probability that banner was main deciding factor.

Now, you run trial campaigns for your product, collect data, and calculate XX base on it, and transfer this XX into your bidding formula. Or trust Google/FB/etc bidding algorithms/ML to do this for you.


Yes, apparently we do.

I'm glad your confidence in your methodology gives you peace of mind.


the results from data analysis give me peace in mind that roas is positive.


enshitification

This also means there is a huge opportunity to disrupt google at this pont.

edit to add reference: https://en.wikipedia.org/wiki/Enshittification


How do you disrupt a monopoly?


I don't think that is the right way to think about google. There is virtually no friction or switching cost for a user to use a different search engine. I think the biggest thing keeping google at the top is that they are the default on mobile phones, and they pay a ton of money to apple to keep it that way.

I also think that disruption will come from something that looks different than google, maybe something like chatgpt, that doesn't appear to be a direct competitor at first but ends up taking a significant bite out of usage of google.


There is a some amount of friction because browsers default to it and user inertia from using it for decades.


And the billions Apple gets paid to pick Google as the default


That's a good point, and I think the big issue here is that Google owns the most popular browser too.


Google's product is not their search engine.


I'm not sure what you mean.


Google sells ad space on certain web properties. Some of those properties are search engines, but that is entirely incidental and not important.

What they actually sell is a B2B SaaS for participating in auctions on those web properties.


Do you know the split between their search ads and the others?


One feature at a time


How this works better for advertisers? I constantly see full-page ads on mobile devices within apps and web sites. I ignore them and click [x] to close. Is this going to count as an impression now and earn revenue?

On this blog post Google says: "Publishers in our ad network are required to adhere to both our AdSense policies and the Better Ads Standards which do not allow practices like pop-ups or interruptive ads that take up the majority of the screen."

But full page ads which interrupt your flow are a standard on AdSense.

Impressions


If you see it, thats already a counted impression.


And that is the point. Suddenly advertised are charged for something I haven't even looked into and had no interest whatsoever (as opposed to clicking on ad to find out more).


Advertisers also have the option of paying for clicks. Paying for impressions is also pretty valuable, this isn't a secret advertisers understand how impressions work.

https://www.bluewinston.com/dictionary/ppi-pay-per-impressio....


Per impression payments are more vulnerable to fraud...

Hide the ad unit in some hidden background layer and you're still gonna get paid for impressions even if no human eyeballs see it...

And you'll get away with it too, because as long as you have millions of web pages and each one only gets a little traffic, there won't be enough statistical power to see that they are underperforming vs just unlucky that none of the first 85 impressions led to a click.

This is already rampant in mobile games.


Doesn't the price per impression still get adjusted based on performance?


Yes, but only when there is enough performance data. The reality of the web is there is a huge long tail of sites that don't get enough impressions to get a good gauge of click through or conversion data - and the fraud makes use of that to make money.


CPM is already the industry standard in most environment, and a cottage industry of analysis/anti-fraud tools have sprung up to fight exactly the kind of hidden ads your describing. (In other words, people are aware of this problem.)


The unintended consequence of paying per click is that it heavily incentives bot fraud, despite what Google might say about 'amazing performance' (Taboola is also a CPC model for context).

This is also another step towards unifying/supplanting the 'Google' ad network and DoubleClick.


If clicks can be botted, why can’t impressions?

What this does kill is a small time publisher opening up their own website while waiting at the Apple Store and clicking some ads they know will pay a lot.

Not that I’d ever do such a thing.


Is it just a question of scale?

You can easily have 100x or 1000x more impressions than clicks.

So detecting fraudulent clicks is much harder than fraudulent impressions, in terms of a proportion of fraud, just because there's so much less of a baseline genuine signal.

E.g. 20 fraudulent clicks out of 40 total clicks per day, who can tell? But 20,000 fraudulent impressions out of 40,000 total, now you can do a lot more pattern recognition to filter them out.


If those 20 fraud clicks didn’t come from a source similar to 2000-20000 pageviews, they should be able to pick out something happening.

Like. You’re right: it’s easier for a small-time fraudster to do cpc fraud, but they should still be able detect suspicious page views, whether there’s a click involved or not. At cpc large scale fraud, there’s a lot of data either way.


Not necessarily at all. The profile of people who click on an ad may be very different from those who just see it.

The point is that small numbers just don't let you make confident conclusions period.


A good example is looking for 'unique devices' that are doing the action.

It is fairly trivial to have 20 unique devices generating 20 clicks, quite a bit more challenging to have 20,000 unique devices generate 20,000 impressions.


It’s a website: you can spoof a lot of details to show uniqueness. Proving you’re not headless is where it gets difficult.


I am really curious about the long term result of the Google and DoubleClick merger. I saw a commentor here a few weeks ago calling it a "reverse-takeover" of Google by DoubleClick. I was around and remember that era fondly so just wondering what was going on behind the scenes and what long-term effects it had.


On one hand Google really wrapped up the whole market quite effectively (especially on the publisher side). On the other hand, I don't know anyone that considers the DoubleClick stack 'nice to use' (best of a bad bunch maybe).

I'm probably naive, but I really believe digital media could be cleaned up significantly simplified if we started over. Programmatic media is massively overcomplicated mess compared to solutions from platforms like Meta.


> to the display industry standard of paying per impression

Huh, who sets the industry standard? Google, google, google and google.


There are other forms of advertising besides online ads. If you're talking traditional print, billboards, an electronic sign in the mall, radio, etc, then you are talking in terms of impressions per dollar.


*per potential impressions per dollar

You can actually use that as a negotiating point. Let's say a new route is opened causing traffic to bypass a normally high traffic point. The advertiser might still price per potential impression per dollar when the reality is significantly different.

Not to nitpick but it's important for people to do their homework on this and realize that those selling the ad space are going to price on the higher end of potential impressions.


So, like Facebook advertising? Honestly I don't trust a single number Google gives me in Google Ads so this probably isn't much of a loss/change.


Regulators should at very least force Google, Facebook and other to use independent corporations for metrics and force them to be publicly available for audit.


The CPM will still factor in click through rates, so it's just a different way of tabulating the same numbers.


The worry is that Google will game auctions and pocket the difference.


Another tweak to the infinite money glitch


Individual web pages will vary a lot in the relative performance of CPM vs CPC, so "not much change" is averaged over many sites. I had an old page of curated links to games that I never updated and the ad targeting was so good the page had a huge click through rate, mostly because the ads were by far the most interesting thing on the page. Another puzzle site I had was the opposite where people could spend 20 minutes to an hour solving a puzzle with just a single page refresh and one ad impression.


I think this a way for them to pay small publishers less.


If Google will pay per view, what incentive will Google have to personalize ads?


google pays money: per view on publisher website

google gets money: per click on publisher website

google is incentivized to maximize click chance per view in this system


Maybe this transition to pay per impressions will help in resolving the issue of unnecessary ad limits


I just hope EU gets properly onto this racket and forces Google to close Adsense.


So will the cost of clicking on ads be reduced?


I wonder if this is related to the current YouTube vs uBlock war?

Could be that a bunch of people have given up on ad-blocking and are just blindly clicking every single ad then immediately closing the new tab to get to the video.

Result on CPC will be ad payments going through the roof, but actual conversions trending down to zero.


Yes, this seems to be a thing: https://adnauseam.io/


lol, ok:

    Safari can’t open the page
    “https://blog.google/products/adsense/evolving-how-publishers-monetize-with-adsense/”.
    The error is: “The URL was blocked by a content blocker”
    (WebKitErrorDomain:104)


I mean, this is the fault of your content blocker being overeager, probably with a *adsense* or *ad* rule in there somewhere


I love how the Facebook Ad Manager and Google Ad dashboards all have a helpful warning telling users to disable adblocking lmao. They know even advertisers don't like ads.


Over aggressive adblocker or custom rule?


Over-aggressive ad blocker.




Join us for AI Startup School this June 16-17 in San Francisco!

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: