October 2021 AdSense Earnings and Observations

[NickMNS]

The countdown to Black Friday, Cyber Monday and the holiday spending season has begun.

[NickMNS]

September was my best month ever, even after the clawback (which is all that really counts). Clawback was <1%.

Also October is starting off strong. But will it last?

[allhearts]

3% clawback for September here, it's more than ever before but still not bad.

[edited by: not2easy at 4:15 pm (utc) on Oct 1, 2021]
[edit reason] splice crumbs cleanup [/edit]

[gatormark]

It looks like the firewall enhancements that I made for September helped. The clawback was still 11%, but much lower than the previous months.

I will continue with further firewall enhancements.

- June clawback 15% (higher than the previous norm of 7%)
- July clawback 24% (new 15-year high)
- August clawback 29% (new 15-year high)
- September clawback 11% (after firewall changes)

[azlinda]

My clawback was 0.03% for September. I can remember many months where the clawback was 30%!

[gatormark]

September ended up being the best month since December of 2017. :-)

[SweetPotato]

8% clawback here.

[gatormark]

October is starting off a little slow.

[allhearts]

Same here @gatormark, but I imagine it's only going to get better.

[puckparches]

The clawback last month was almost half from previous months, which I tough was for an increase in IP numbers blocked on my firewall but looks like everyone's clawback was lower last month.

September ended up being the best month since December of 2017. :-).

For me the best month ever.

October is starting off a little slow.

Same here

[IndyMattR]

Sept was good but now seeing big drops for October here.

[delorean]

How to see clawback?

[NickMNS]

Select "payments" from the left sidebar menu (near the end), then "payment info", then in the "transactions" box click "view transactions". You should then see a listing of the transactions including your previous month's revenue and if applicable (almost certainly is!) "invalid clicks" which is deducted from your revenue. This deduction is what we here refer to as "clawback", because Adsense takes the money back with it's evil claws!.

[NickMNS]

Side note I was joking about "AdSense's evil claws" being the source of the name. "Clawback" is an actual term used in finance and banking:
[corporatefinanceinstitute.com...]

[SweetPotato]

Page RPM plummeted Since October 1st.
My Impressions are stable.
Earnings -20% vs previous 7 days.

Anybody seeing something similar with RPM?

[allhearts]

Yes @SweetPotato, and from the looks of the previous posts, we are not alone.

[gatormark]

Page RPM plummeted Since October 1st.
My Impressions are stable.
Earnings -20% vs previous 7 days.

Anybody seeing something similar with RPM?

October started off a little slow, but the past three days I am seeing a Page RPM increase.

Today and Monday have been two of the best days, revenue-wise, in the past few years.

In my niche Sept-Dec and Feb-May are the money months.

[CommandDork]

Clawback for me, after labeling the ads per Adsense support's suggestion, was better - with 25.6% in August now down to 13.7% in September (albeit earnings were lower overall in Sept when compared to Aug). I went a step further and offered challenges in Cloudflare to trouble-making countries.

Support did say to give it a few months to see how things go - so hoping October can get down to single digit clawbacks. But definitely an improvement for me. October's haul is looking good for me right now but it's way to early to pop the cork.

[gatormark]

October's haul is looking good for me right now, but it's way too early to pop the cork.

@CommandDork

Me too. October's numbers, and especially today's numbers, are back to my best days of 2015-2017. However, I'm not counting my chickens before they hatch. Still, September was the best month since December of 2017, so this does bode well. I won't get too excited until November 1 at around 1 PM.

[MayankParmar]

Evolving the AdSense ecosystem: Our move to a first-price auction

tldr; no changes for publishers.

[CommandDork]

Ya, the first-price auction is an interesting change for the program if I understand it right: might mean good things, bad things, or nothing at all for us publishers.

Probably need about six months of data to be sure either way.

[RhinoFish]

As the ad agency on the other side of the AdSense transaction, the switch to 1st price auction very likely means we will turn bids down.

In our most relevant targeting, where we really fit, the 2nd price auction rewarded us richly.
Seeing it begin to be removed in this one part of AdSense is not giving us a warm fuzzy feeling.

Some details here...
[searchengineland.com...]

And a whopper of a quote...
~~~~~~~~~~~~~~~~~~
This will simplify things for advertisers, Google says. “On display ad selling platforms, a first-price auction simplifies the buying experience for advertisers because the final price reflects the winner’s bid,” Google said on its FAQ page about AdSense moving to a first-price auction.
~~~~~~~~~~~~~~~~~~

Zoinks, "simplifies the buying experience", ouch, that stung.

[NickMNS]

I have got to say that I am utterly confused by everyone's reaction to this change (*well not really). Back in January, I had a long drawn out discussion about this very topic, where a one users suggested that Google was screwing it's AdSense publishers by continuing to use the First Price auction model, as opposed to Ad-Mod and Ad-Manager which were already using second-price. To this I responded that it was a moot point as there is no difference for the publishers. The general sentiment here was, that I was wrong and that second price was better because you could have bidding wars (which you can't).

Well folks, Google heard the whines, and second price is here, and now everyone is claiming that the change will be bad for publishers. But the bottom line is as claimed by Google, it will not change anything for publishers. Any swings in price that may result from this will be dwarfed by normal market fluctuation that we see everyday. Ad market are efficient with many knowledgeable players, the auctions will close at the market price like they do now.

Assuming, that prices do fall, the likely outcome will be that buyers will maintain the same budgets. They will then be able to bid on more auctions creating an increase in demand and an increase in price, pushing the market right back to where it started. Again no change.

*well not really:
Not really because the general sentiment around here is that no matter what Google/Adsense does or announces, it's bad for webmasters/publishers. This is a perfect example of that. In January it was framed as Google not switching is bad for publishers, and now that we have forgotten what was said in January and that Google announced the switch, it is now framed as bad for publishers.

[robzilla]

I started a 90-day experiment 24+ hours (and 25,000 pageviews) ago, but there are still no stats available:

"Experiment is still collecting data. Results will show here when they're ready"

How long does that generally take?

[ember]

We just have to wait to see what the auction bidding change means, if anything (I doubt much will change). As always, diversifying revenue streams is a good idea in case Adsense drops, stops, goes belly up, etc.

[robzilla]

I started a 90-day experiment 24+ hours (and 25,000 pageviews) ago, but there are still no stats available:

"Experiment is still collecting data. Results will show here when they're ready"

How long does that generally take?

Never mind, looks like it's now reached statistical significance. Oddly enough, revenue increased quite a bit during the experiment but the original version is declared a winner over the variation, by a hefty margin too. It should be the other way around. As if they swapped the original and the variation somehow.

[NickMNS]

Oddly enough, revenue increased quite a bit during the experiment but the original version is declared a winner over the variation, by a hefty margin too.

Why is that odd, it is the reason for the experiment. Revenue varies wildly with AdSense and correlation does not equal causation. The experiment allows you control for these random changes in revenue.

As if they swapped the original and the variation somehow.

During the experiment they show both versions. Over some large number of impressions they show the original to some users and the new version to others and then compare the results.

[robzilla]

Revenue varies wildly with AdSense and correlation does not equal causation.

True, generally, but not this wildly: the difference was 65%. But this was attributed to the original instead of the variant, whereas the variant was much more aggressive and was bound the outperform the original. The experiment enabled auto ads (I was curious) and I can tell from the reports that the vignette ads that ran had RPMs nearly 20x higher than my usual ads.

The experiment ended and I disabled auto ads again, those vignettes are just too obtrusive. But I'm running the same experiment on another site, will wait to see what happens there.

[gatormark]

Page RPM has dropped considerably today, -25%. Is anyone else experiencing this drop today?

[Mentat]

Traffic drop? Yeah…
Discovery changes again.
As revenue, this month is not great, but not terrible.