Adsense's new Ad Balance slider

[Evan Salamanca]

Anyone notice this yet? You can change the balance of ads that show kind of like FBAN's balancer on steroids. It even shows you the percentage of ads that show versus percentage of potential income. If it weren't for the fact that the ads units show as blank rather than collapsing entirely, this would sound like the best thing ever. Although this creates huge potential when combined with good backup ads.

[EditorialGuy]

I don't think that contextual vs personalized ads matter in this case. I think the issue is a question of a sites ability to attract high quality ads.

I think that, quite often, it's less about the site or ad quality than about the audience. It would seem to me that a site with a global audience is likely to have more low-paying impressions and clicks than a site with a primarily U.S. audience (all other things being equal), simply because there's likely to be less competition for clicks and impressions in smaller markets.

So (for example) a travel-information site might be better off devoting low-value ad impressions in smaller markets to house ads for a hotel affiliate program, which generate the same commission on bookings from Ljubljana as on bookings from Los Angeles or London.

[NickMNS]

@EditorialGuy I agree with that. Clearly geography plays a big role. Magician described the problem on page1. Blocking low revenue ads could end up blocking whole geographic areas that have low payouts (eg: india vs US). It is possible that Adsense accounts for this but I would not bet on it.

[IanTurner]

I have large numbers of South American traffic on my spanish site and am seeing a revenue drop - so you may be right. I am going to run Ad Balance at least into next week to see if things improve.

I am seeing some indications that bounce rate is improving as a result of showing less ads - but this is not across the board.

[Lagonda]

Feedback post.

Adjusted to 51% ads, 98% revenue (yes, it's THAT bad).

Revenue down by 10%.

Of course this isn't a one factor result, we have to consider there are plenty of factors involved but, nevertheless, here's some feedback.

AdSense team, if you're reading this: I welcome this setting, thanks.
Now, if only the empty ad space could collapse, that would be great.
If not possible, a matched content unit as a backup ad would work too. ;)

[james_09]

I can get my slider down to 55% ads and still retain 100% earnings. So I guess my sites must be attracting a lot of low quality/spammy ads.

I have roughly 50,000 ad impressions a day from Adsense. Is google really saying that 22,500 ads a day appearing on my sites are earning me zero?

Why are these ads even allowed to show up... Surely you have to bid something higher than zero to appear in the first place. Unless it's saying that 22,500 ads a day are so bad they have zero chance of being clicked.

If I put it down to 1% it is telling me I can still earn 33% of my earnings!

How accurate are estimated earnings? I think I must be missing something...

[breeks]

My sites 90% ads = 100% earnings.

So 10% are bottom feeder ads. Seems like there are more than that.

[NickMNS]

@James to understand what is going on you can look at the revenue profile report in Adsense [Performance Reports -> Revenue Profile].

The report shows a curve, where the left most points show the average rpm of all the ad request, as the you move to the right you trim off the least profitable ad impressions. If you go to 50% the average rpm shown will be for the top 50% highest paying ads. This is the data that ad-balancer is based on. To check it you can do some math:

1- Take the avg_rpm @100% mark * total impressions/1000 for the time period (eg 28days) this should equal the your revenue of the period.
2- Take the avg_rpm @50% mark (or any other suitable point on the curve) * total impressions/1000 for the time period (eg 28days) this should equal the revenue provided by the top paying ads for that period.
Now take the difference of the two numbers and divide it by the total revenue and you should end up with the same percentage number as the ad-balancer.
You can also take these numbers and figure out the RPM earned on the bottom 50%. In my case I calculated it on the bottom 30%, and it gave me an RPM of $0.10.

So now if you look at the shape of the curve it can give you some idea of the strategy to take.
A largely flat and horizontal curve means that you will benefit very little from the ad-balancer.
If you have a curve where the left tails is flat and at some point it begins to curve up steeply, then you would want to select the point near to where the curve starts as your ad-balance cut-off.
If that curve happens very far to the right, then again you may want to question the use of the ad-balancer.

I have roughly 50,000 ad impressions a day from Adsense. Is google really saying that 22,500 ads a day appearing on my sites are earning me zero?

Effectively yes. Actually you are being paid just very little. See my calc above, I earn an rpm $0.10, that means I need to get 10,000 of these impressions to make 1$.

Why are these ads even allowed to show up... Surely you have to bid something higher than zero to appear in the first place. Unless it's saying that 22,500 ads a day are so bad they have zero chance of being clicked.

Yes advertisers bid, but most bid based on cost per click. So an advertiser may win the auction but if your user doesn't click the ad you make only fractions of pennies (for the impression only). What is worse, is that spammy advertisers may be bidding and winning on your ad space thus displacing legitimate advertisers that shows ads that have a higher probability of clicking. I don't think that the ad-balancer fixes this situation directly, as the legit ads (at the current bid price) may be blocked as much as the spam ads. But advertisers may be more willing to bid higher for your site if they see fewer spammy ads.

[IanTurner]

Thank you NickMNS, looking at my revenue profile there is a lot of flat at the left.

[Rasputin]

One aspect that isn't clear...on long pages with three ads (one near the top, one near the middle and one near the bottom) it can be the first or second that are not shown because of low earning potential while an adsense ad still shows further down...
But it is clear from years of stats that the best performing ads are the highest in the page, as you would expect.
So on what basis would the first ad be removed as 'not contributing to earnings' while later ads are still shown?

[ember]

I'm at 90% and 100%. EPC is about the same but the CTR is up, RPM is up and earnings are up. Almost feels like the old days.

So on what basis would the first ad be removed as 'not contributing to earnings' while later ads are still shown?

I have long pages, too, but it is usually the ad at the bottom that is not showing. This is a new feature, I think, and maybe still has bugs.

[NickMNS]

@ember

CTR is up, RPM is up

This is the hoped for outcome of the ad-balancer, if you eliminate the impressions that are not being clicked then the impression gets smaller.

Impressions is the denominator for both CTR and RPM. If the denominator gets smaller and the numerator stays the same this results in a higher value. But don't be fooled, the numerator has not changed and it is only the numerator that drive revenues (eg: actual clicks and actual revenues). So be weary when comparing the figures with ad-balancer switched on to those historical pre ad-balancer figures.

@rasputin, As for long pages issue, I assume that the it is based on the outcome of the auction. If for whatever reason the top ad get a low price at auction and the lower ones get high prices, then it would be expected that the top ad is removed and not the lower ones.

[anefarious1]

I'm still confused. If the ad unit doesn't display (because I'm using this feature).. does it have to be a white space? Can a backup ad be shown if I use the Ad Balance Slider? Definitive answer?

[Dimitri]

@anefarious1 : backup ads are showing.

[EditorialGuy]

Sure, backup ads will display if they're implemented correctly. Just allow a little time for them to show up. (Whenever I've added backup ads to an ad unit, it's taken an hour or two for the AdSense servers to roll out the change.)

[anefarious1]

Google says that a white space will display but I don't believe them. They say nothing about the ability to use a backup ad when using this so-called slider.

[Dimitri]

@anefarious1, I answered you , because I am myself, using he Ad Balance feature, so I confirm you that backup ads are show. You can see in my different posts, that I mentioned this. On the first page of this topic, I did pointed that Google claims it will show an empty space, and motorhaven confirmed that his backup ads were showing. So I started using it, and I can confirm you that the backup ads are showing. Now as mentioned EditorialGuy, it can take a few hours, before your backup ads start showing (this happened to me) , and also, all this is considering that you are correctly implementing your ads. You have to set up an URL and Google will display the content of this page within an iframe.

[Rasputin]

It's not a question of believing or not believing - try it and you will see that they do show (possibly after a couple of hours delay).

[anefarious1]

Yes I believe you. What do you mean about the URL? Isn't it a 3rd party ad unit code.. not just a URL.

[anefarious1]

Consider that Google's goal with this ad slider is to reduce the number of ads served. The backup would seem to be counter productive to this end. When backups are served it works more like a passback code (many other networks have long had pass-back ability). Interestingly, Adsense has always had backup ads but they almost never get served because the fill rate has always been something like 99.8%

[NickMNS]

@anefarious1 the backup ads are served through an iframe thus they cannot be targeted to the user. The ad are always served to the same html page (the one you create to show to hold the ad) and that page will then be shown to the user in a iframe. Passback code is available in Adsense by enbaling third party ads. The goal of the Ad-Balancer is not limit the ad served per se, but rather eliminate ads that are poor quality. Replacing poor quality ads with more poor quality (low quality ads tend to appear when untargeted) from another ad networks defeats the purposes.

If you have contextual ads such as relevant affiliate links then this could be a good solution.

[ember]

I'm seeing CTRs and RPMs that I haven't seen in years. It's all good. Almost too good.

[IanTurner]

@ember0 - some questions

What settings did you use?
How long did it take for the new CTRs and RPMs to show up
Are you talking about page RPM or impression RPM?
And what is the overall effect on your income? (Has the income at least stayed at the same level? - mine went down by about 15%)

[EditorialGuy]

Google says that a white space will display but I don't believe them. They say nothing about the ability to use a backup ad when using this so-called slider.

It isn't a "so-called slider," it's a slider.

And if you look at your ad units under "My ads," you'll see options for displaying white space, a block of color, or backup ads when AdSense ads aren't being served.

[edited by: EditorialGuy at 6:29 pm (utc) on Jan 15, 2017]

[NickMNS]

@ember did you see my comment to you above, don't get too exited about the increased CTR and RPM, this is a side effect of the ad-balancer.

Since it is no longer showing the ads that didn't get clicked it reduces these impressions so the CTR/RPM number go up. This does not necessarily mean that you are getting more clicks or getting paid more. See my comment above for a better explanation.

[ember]

Settings are 90% and 100%. Started seeing an uptick the day after I applied the balancer. Things really jumped yesterday and today, so about 3 or 4 days for it to really kick in. Both page and impressions RPM are up. Income is up 40% over a similar period before the balancer. EPC is up, too, but I think that is because we're in the middle of January when epc normally rises anyway.

I'm not excited. I'm a little worried.

[EditorialGuy]

wonder which can on impact , this feature will have on Advertisers. May be it's a trick to force their to increase their bids. It seems Google uses the argument of the "user experience" to enforce lot of things...

I wouldn't call it a "trick." A better word would be "strategy." And it isn't a strategy to force advertisers to increase their bids, because the advertisers who are being filtered out by the ad slider are likely to be bottom feeders who can't afford to increase their bids.

IMO, the real purpose of the "Ad Balance" slider is to make it harder for those bottom-feeding advertisers to get exposure. This has at least two benefits that I can see:

1) Publishers are happier, because they can allocate ad impressions more effectively while improving their user experience.

2) Fewer junk ads = less ad blindness.

[ranka]

TO PREVIOUS POSTER: 2) Fewer junk ads = less ad blindness.

Junk ads and ad blindness - they are not related, actually vice versa. Junk ads are usually spam, which are more attractive than usual ads. E.g. Doctors Just Can't Believe How She Lost 15 KG in 3 Days or She Lost So Much Weight That Even Her Mother Did Not Recognize Her . So they are seen and clicked more than quality ads. It's good for you - you are getting money by good CTR, but there is one big problem.

The problem is that junk ads are bad for user experience. If visitor recognize it's a spam, it lowers your website quality. If visitor do not recognize or/and click on these ads, then usually she is redirected to spam site. No-one do not want visit sites with full of lies or sites which redirect them through the ads to spam sites with lots of popups, wrong information etc.

But overall - do not block spam because Adsense is the largest ads network and millions of sites show these ads anyway. You are not losing customers and visitors because of that. You are not exception, you are rule and so long as Adsense showing these ads over the network, you should keep them.

Also I think that only few percent of Adsense users have backup ads which have better RPM than Adsense lower quality ads. If you do not have this, keep the slider 100% fill-rate, if you have (or experiment), then it's worth to try.

[Dimitri]

@EditorialGuy, my words were slightly ironical :-) even if I think it can put the pressure on some kind of advertisers.

@ranka, if I don't make mistake, there is something called "smart pricing", so I assume that if people are clicking on spam ads, this can downgrade the quality score of your Adsense account, since they will not stay long on the advertiser's site, or do any action there. Isn't it ?

[EditorialGuy]

Junk ads and ad blindness - they are not related, actually vice versa. Junk ads are usually spam, which are more attractive than usual ads.

I can't imagine why anyone would think a generic ad for "downloads" is more attractive than a relevant, targeted ad (or more likely to get clicked). Also, CPC ads are only one piece of the pie. On a strong niche site that reaches a desirable audience, CPM ads can be an important source of revenue.

Also I think that only few percent of Adsense users have backup ads which have better RPM than Adsense lower quality ads.

In our case, the backup ads are "house ads" with affiliate links or house ads that promote our information site's affiliate-related content. Such impressions are a far better use of our screen real estate than cheap generic AdSense ads are.

From my point of view, the "Ad Balance" slider is a win-win:

1) It lets us use the 300 x 600 ad space in our right margin more profitably.

2) By increasing the percentage of highly relevant, targeted ads, It improves the likelihood that ads are actually useful to our readers (which is another way of saying that it improves the user experience).

As always, YMMV.

[EditorialGuy]

Something else to keep in mind: Your "backup ads" don't need to be conventional ads. They can be nearly anything that fits within the ad unit's dimensions.

For example, I've created a backup house ad for one of our AdSense units in the form of a simple HTML page with a photo, a block of text, and a clickable "call to action" line. It's essentially a micro-article that provides useful information while supporting the affiliate links above it, all served up in a 300 x 600 space.