January 2021 AdSense Earnings and Observations

A very pleasant surprise this morning.

[azlinda]

---

System: The following 11 messages were cut out of thread at: https://www.webmasterworld.com/google_adsense/5018422.htm [webmasterworld.com] by martinibuster - 3:29 am on Jan 2, 2021 (utc -5)

---

A very pleasant surprise this morning. After endless months of 30% clawbacks, it was only a little under 2% for December.

[Niresh12495]

That should do it. Yes still it be considered responsive.

If you want to use fixed size units the following both methods should help.

If it is sync ads you might try something like the following.

[guest.blog...]

Or Async

Use madgex solution, the best of all, little complicated but best. (You should remove comments while using following method)

[github.com...]

[Niresh12495]

@Lisa01 About your matched contents, there is some option in Adsense dashboard where you could manually choose some pages to be shown in matched content recommendations.

[allhearts]

That's a little complicated for me @Niresh12495,
the method I posted would not work?

[Niresh12495]

Yes that should improve earnings.
For desktop use 970x250 or 728x90
For mobile use 300x250 size

[allhearts]

Thank you, I really appreciate your help with this, I'm not really a tech guy.

I will keep you posted.

Thanks again.

[Niresh12495]

You’re welcome, a little tweak in adcodes could make a bigger difference in earnings.

[Lisa01]

I can create matched ad units. It's just that I'm baffled by the message it is showing

[allhearts]

@Niresh12495

Good to know, thanks again.

[NickMNS]

Should it improve or still be considered responsive?

Should it improve, not likely but no way to know for sure. Theoretically, it should perform worse, as it restricts the ad-unit to a specific size. Some ads themselves may be designed to adapt, so it isn't completely limiting but this may skew the layouts makings them appear odd to the user (I have never seen this occur but it is in the realm of possible). And "Yes" it will be considered responsive.

AdSense is a game of supply and demand, so the more you restrict your ability to take the demand, the lower the price you receive. Restricting ads to non-responsive should again in theory provide lower revenue as your are restricting all the responsive demand. The reason that some see the opposite, is likely that their implementation of responsive ads is poor, causing ads to be misplaced or overstep the area laid out for them on the page, thus resulting in fewer clicks.

Using media queries as you propose is a good idea as it can set bounds on the ads displayed, but I would make it less restrictive by setting a max-width and max-height instead of width and height. You also want to consider setting a min-height so you don't get filled with 50px height mobile banners.

[Niresh12495]

Resposive ads doesn’t bring much revenue. Ever since the beginning i have noticed fixed size brings much revenue. You can look into creative sizes from reports section of Adsense report, to see which ad size performs well. If adsense is a game of supply and demand, websites which got most ads should win, it doesn’t. After few optimization I’m getting around 2$ RPM, This January. At the beginning rpm was around 0.5$.
Its all about optimization.

[NickMNS]

Resposive ads doesn’t bring much revenue.

Well I use responsive ads exclusively and my revenue is better than what you report by a good margin, mind you revenue may not the be the best measure as it heavily dependent on your niche.

You can look into creative sizes from reports section of Adsense report, to see which ad size performs well.

Unfortunately it is not that simple, for example when I look at my report all I see is how great my responsive ads are performing, but I cannot see how much better the non-responsive ads that I did not use would have performed. You could show responsive ads on some pages and non-responsive on others, but again ad performance tends to be page dependent, so the result may not be accurate. The only way to test it, would be using an A-B test. Unfortunately Adsense removed the ability to A-B test any ads other that Auto-Ads. So you would need to set that up yourself.

If adsense is a game of supply and demand, websites which got most ads should win, it doesn’t.

I'm not sure I understand that statement, but the volume of ad-fills is not a measure of success, in fact one needs to be performing pretty badly to get low fill rates, because there are always advertisers willing to buy un-filled inventory on the cheap. See the comments about ear-wax ads earlier in this thread, as proof of this.

AdSense works by auction, your impressions consist of your inventory (not the number of pages of your website), that is the supply. Ad buyers bidding on your impressions are the demand. Getting buyers to increase their bids for your impression is "success". If there are only 10 buyers and you have only 1 impression, those buyers are likely willing to pay a lot for that impression maybe 10$ for a click, whereas if you suddenly have 10 impressions, then those 10 buyers can bid a penny and get filled, that is supply and demand.

Now if you have that 1 impression but you limit the ads that can appear to non-responsive, then there may only be one or two bidders willing to buy those ads, you have restricted demand and thus you will likely get a much lower price for that impression.

My example is over simplified, because your impression is 1 of millions of simultaneous impressions, and buyers are likely willing to keep buying impression as they become available, but the dynamics remain the same.

Restricting demand lowers prices.

Blocking ear-wax ads, is restricting demand and thus lowers prices, but there are situations when taking a lower price is beneficial in the long run. If users see spammy ads on your site, then they are not likely going to return, this will lower impressions and restrict your supply and lower your revenue.

The key is understanding dynamics of the market that you are operating in, such that you can make an informed decision as to what is right for your specific situation. If you have the means and the volume to warrant it, you can always A-B test to make the best choice, but I think for most us this will likely cost more than it will return.

[gatormark]

My revenue has always been lower when I used responsive ads exclusively. That’s just my experience.

[Niresh12495]

So am i @Gatormark, Also ezoic adtester is famous for increasing revenue over time by testing adsizes, If it isn’t necessary it wouldn’t exist, if a single ad-size could increase revenue, everyone should get decent amount of earning, they don’t.

Google only cares about advertisers not publishers. Even adsense runs in second price auction. Google never wants to pay publishers more.

This is from Ezoic.

The impact of ad size by visitor location.

300x250 Higher CPMs in USA
300×250 ads are very versatile and typically see a large number of advertisers bidding on them in the USA.

250x360 Higher CPMs in Sweden
Uncommon in most of the world, 250×360 ad sizes are very common and highly competitive among Swedish advertisers.

[Niresh12495]

I personally does not use Ezoic as they want to control websites from DNS level.

[ember]

I tried Ezoic and my revenue tanked.

[Niresh12495]

It depends on the niche site speed and all. Ezoic slows down site according to most users.

[MayankParmar]

Google just sent another email to remind us about link ads retreiremnt and clarify the date March 10th 2021 :(

[Niresh12495]

I saw the mail too, i was very surprised by the discussion of people saying they earn upto 55% of revenue from Link ads.
[webmasterworld.com...]

[Lisa01]

I don't know what to do now. I earn the most from link ads

[Lisa01]

Even google must be knowing that link ads work well for most publishers.. Don't know why they're doing this. Looks like one day they'll discontinue everything, leaving responsive ads.

[Niresh12495]

Thats what I’m afraid of.

[NickMNS]

Even google must be knowing that link ads work well for most publishers.

That may be so, but it is not publishers that pay the bills, it is advertisers and it must not have been working well for them.

[Niresh12495]

Google always work in favor of advertisers, like Keeping adsense in Second Price Auction.

[NickMNS]

like Keeping adsense in Second Price Auction.

How exactly is that a bad thing for publishers? Second price auctions allows the buyer to place their bid at the absolute maximum they are willing to bid, knowing that the price filled will be 1 penny more than the next highest bidder. Now the other bidders are likely doing the same, the result is a maximization of the final price.

Here is an example, with only two auction participants
So bidder one bids 5$, knowing that the price will be lower but to be sure to win the auction adds a bit of margin. Now bidder two has the same expectation, she figures the price will be about $1.50 but to sure adds some margin and bids only 2$, the price is then $2.01. Now you maybe thinking, "That sucks, bidder one bid 5$ I should get the $2.99 too". But that is not how it works.

In first price bidding, buyers need to place bids at the minimum bid that they think will win the option, the net result will most likely be a lower price.
So buyer one would likely bid $1, then buyer two bids $1.50, buyer 1 counters at $1.75, now buyer two sees prices climbing and thinks, "I only wanted to pay $1.50 I better stop here". Thus the auction closes at $1.75. The net result is that compared to second price bidding the auction price ended lower. Yes, in this case the buyer one will pay $0.25 more that the previous bid, but that bid is $0.50 lower than it would have been in the second price auction.

Now ad buyers are an intelligent bunch with deep pockets, so they likely have derived suitable algo's to adjust for the specific auction type that they are bidding on and ultimately the end result will that there is little to no difference with either auction type.

Just a quick aside, the most likely reason that second price auction was adopted was for speed. A first price auction requires time for auction participants to place counter offers, whereas the second price auction only requires a single bid from each participant.

[Niresh12495]

If second price auction was best then why Google AD Exchange shifted to First Price Auction?

All ad industries are moving towards header bidding, on first price auction aka Unified Auction. That was inevitable for Google to avoid. Still they control Adsense publishers and with their dominance they force them to be in Second Price Auction. They don’t allow them to join ADX.

In the above example what you have given if it is a CPC bid, On frist price auction publisher will be getting 5$, would any publishers here go for 2.01$ or 5$?

[theversed.com...]

Expensive Ads, Expensive Payout for Publishers.

[ember]

Always remember the Golden Rule: He with the gold rules. Advertisers pay the bills. Google will always try to please them. Publishers are kind of an afterthought. If you want to know why something is happening or why it is the way it is, follow the money. As NickMNS says, the link ads must not be working for advertisers.

[Niresh12495]

As a publisher i just want to maximize my revenue. Thats all.

[NickMNS]

In the above example what you have given if it is a CPC bid, On frist price auction publisher will be getting 5$, would any publishers here go for 2.01$ or 5$?

No! Not at all. I think this is where your confusion lies. The 5$ is simply a number that goes nowhere, to get to $5 there would have had to have been a bidder willing to pay $4.99. You can't simply apply the numbers from the second price auction to a first price scenario.

Let me explain it differently, in a first price scenario rational bidders must always bid the lowest price at which they expect to win the auction (bias to lower price). Because (using numbers from my example in my previous post) if a bidder bids 5$ and the next highest bidder bids 2$, one just over paid by 3$, thus to avoid the scenario one bids low.

In a second price auction the rational bidder must bid the highest price they are willing to accept, knowing that there is a high probability that that price will not be reached. (high price bias). Here if the bidder bids 1.50$ (the low bias price) but your are willing to pay up to 5$, then another bidder bids 2$ you lose the auction despite the price being well below your threshold.

The bottom line is that bidders will bid differently in both situations, so no one was going to get 5$ ever. Shifting to first price bidding will result in lower prices for the bidders and the shift to this type of auction on platform such as ADX is exactly to benefit the advertisers, not the publishers.

[NickMNS]

@Niresh
I should also add that the author of the article you linked to also made the same mistake as you in her example. But, the article goes on to point out that the programmatic tools used need to be updated for use with first price auctions and the reason is that they need to handle the "low price bias".

[NickMNS]

And now a correction...

My comment:

Just a quick aside, the most likely reason that second price auction was adopted was for speed. A first price auction requires time for auction participants to place counter offers, whereas the second price auction only requires a single bid from each participant.

This is incorrect as I mixed up two types of auction. In a first price auction only one bid is ever made. It is in "Open ascending-bid auctions" that subsequent bids are made. That said, it is all the more reason to ensure that you don't bid too high.

For information about auction theory:
[en.wikipedia.org...]

Note that this years Nobel prize in economics went to two researches in Auction Theory, Robert B. Wilson and Paul Milgrom.