Fun fact – 99% of experts will answer online advertising questions with “it depends”. But not here at Online Advertising Answers! Mostly because I hate giving that answer. It may be right most of the time but it’s not useful to hear.
So keep reading for practical advice that only uses the word ‘depends’ when I can’t avoid it! In this issue, I explain AdSense CPC, how to breathe new life into an old Facebook campaign, and examine a digital marketing riddle – why more impressions can mean less revenue.
You asked, I answered.
Let’s dig in.
Have a digital marketing question?
This month’s questions are:
- How is the Google AdSense CPC calculated?
- Can I extend the life of an FB campaign with new ads?
- Why is it that sometimes more ad impressions means less revenue?
[Click these links to skip down to the question]
Q: “How is Google AdSense CPC calculated?
If a website has 50,000 page views per day, 1% CTR with $0.25 per click and displays 4 ad units on each page. What is the right way to calculate the CPC Adsense revenue of that website?
While calculating the CPC should I multiply the number of page views by the number of ad units? Or only look at ad units?“
&
A: “CPC stands for Cost Per Click so in your example, the CPC is $0.25 as that is the amount paid per click.
CPC on AdSense is calculated as the total amount you have earned divided by the total number of clicks.
Click to enlargeCPC = Cost ÷ Clicks
To improve your CPC
You don’t need to use your CTR, Page Views or number of ad units to calculate your CPC. However:
- A higher CTR will lead to AdSense sending you better ads, which will pay more (which will improve your CPC if people click on them).
- Having more Page Views will probably increase the number of clicks you receive, which will increase your revenue (although not necessarily your CPC).
- The number of ad units on a page can affect your CPC. Having too many on one page and your CPC will go down (as people usually only click on one ad per page at most).
In your example you talk about a website with 4 ad units per page – I would not recommend having more than 3 ideally. This is so that you don’t annoy your users – which in the end will make you more money (through return visits) than simply having an extra ad on a page.
On top of that, you should try and make sure your ads are all “above the fold” if possible. This means they should be on screen when a page first loads.
If you do that then you may get a little bit of money every time a page loads regardless of whether someone clicks. This is because lots of advertisers are moving towards a system where they only pay if their ads are actually seen. Having your ads above the fold will help make that happen.
To summarise
- To work out CPC, divide all your revenue by your number of clicks
- To increase your CPC don’t have too many ads on any page, and put them at the top of the page.“
[This question was originally on Quora]
Q: “Can I extend the life of an FB campaign with new ads?
Hey there,
I know a campaign can eventually die out once it’s reached its current audience enough times and ad fatigue sets in.
I’m just wondering, if you add a new piece of content to an exhausted campaign / ad set, will it come alive again and start sharing that new creative to people? Or will it ignore the new creative and still see the audience as exhausted?
Hope that’s clear enough, and thanks in advance!“
&
A: “You shouldn’t wait until the audience is completely exhausted – but yes changing up creative can definitely breathe new life into a campaign.
If you’ve completely burnt out an audience (e.g. few-to-no conversions coming from them anymore), then you probably need a significantly different creative to jump-start them (and Facebook Ads) again.
You should consider testing small creative changes to ads (testing new text or images) as part of your standard optimisation process anyway. This means every couple of weeks (depending on your level of spend) that you cut the worst ad and try a new variation.
However, if the campaign is barely serving anymore then a big change of creative is probably your only hope to revive it quickly.
In this case, start a new ad set with it (if there is some reason you need to stick with the original campaign) or consider starting a whole new campaign if the content is different enough. It can be that FB still considers the audience exhausted if you stick it in the same ad set or campaign, so breaking it out will help counteract that.
Personally what I like to do when I’ve had a great campaign that is now flagging is just give the whole thing a 3-6 month rest. I then start a new campaign with different targeting (even if it’s only slightly different) and see if I can improve on the results. When the new one starts flagging, I either start another new one or test switching the old one back on again.
I do this because if you keep messing with what was once a successful campaign, you might lose what made it great in the first place. So instead, switch it off and try something else. Leave your once-great campaign frozen in time so that you can give it another shot in the future. It may not do quite as well next time around, but in my experience, they still often come back pretty strong.“
[This question was originally on Reddit]
Q: “Why is it that sometimes more ad impressions means less revenue?
How do ad impressions work? How come sometimes more impressions get less revenue than more impressions?“
&
A: “An ad impression simply means that one ad was loaded one time. The way they work is that for each ad slot on a page, one ad impression is counted.
This means that one page view = many ad impressions (as there are usually multiple ads on each webpage).
So when a webpage loads, it also loads all the ad slots on a page, and each ad slot will hold a bit of code from an ad platform. That bit of code will tell the ad platform about the ad slot that is being loaded – and it more than likely will tell them a bit about the person loading the webpage too.
The ad platform will then take everything it knows about the ad slot (e.g. what page it is on, what size the ad slot is etc), and the person (e.g. what country they are in, what browser they are using etc) and then auction off that ad impression.
This happens surprisingly fast (less than a second generally) and the auction will include a surprising amount of advertisers (possibly thousands). The advertisers will see ad impressions that come up for auction and bid on them based on how likely they think they are to get what they want from them.
This is a simplified description of what happens continuously all across the internet every single second of every day. The more advertisers that want an ad impression, the higher the (potential) price. Potential is the key word here though. Let me explain:
Example of a simplified ad auction
Imagine a pet supply company in Spain (let’s call it Petso) sees an ad impression come up for a webpage about puppies. The ad slot is 300×250 in size and it’s 7:30 pm. The person viewing that webpage has also looked on Amazon for dog bowls in the last week.
Petso is looking for sales and it knows that this ad size has worked well at this time of night before, especially on this website. The person viewing the webpage appears to be in the market for dog-related items so it’s likely the ad will work.
Petso makes a high bid on the ad impression of $3.50. It will pay that if the ad impression results in a sale. It sends this bid to the ad platform the puppy webpage uses.
The ad platform (let’s call it Adso) weighs that bid against others it has received. Adso knows that Petso is a pet company, that the webpage is about pets, and that the user has looked at dog bowls recently. So it guesses that there is a 50% chance that giving them the ad impression will result in a sale for Petso (and therefore a payout to Adso). Therefore it internally thinks of the bid as $1.75 (because $3.50 x 50% is $1.75).
Adso has many other advertisers trying to get that ad impression too. None of them are pet companies though, and none of them are bidding nearly so high. So Petso wins the auction and their ad is served.
However, the person on the website does not click on the ad. They do however stay on the website and browse more pages creating more ad impressions. Petso keeps bidding, but each time they bid a bit less as they are less confident of getting a sale.
Eventually, the person notices the ad and remembers they still need a dog bowl. So they click the ad and make a purchase. Petso pays Adso who pay the website (about 50% of what Petso pays them).
So… when do websites get paid?
Continuing with the above example, if the person on the site had never noticed Petso’s ad then Petso would have eventually stopped showing its ads to this person, no matter how many pages they went to. These ad impressions would therefore generate no revenue.
Similarly, if that person had clicked the ad, but not bought anything this would not have brought any money in. The advertiser (Petso) would only pay out for sales, so anything else wouldn’t count – and so again the ad impressions would generate no income.
Advertisers can buy ads for lots of different reasons (clicks, engagements, conversions etc) so this happens all the time.
This is why you shouldn’t think about individual ad impressions too much. A single ad impression is unlikely to earn any revenue. However, when you have thousands and thousands of them that equation changes.
When something is improbable but has essentially infinite chances, it becomes probable. Think about bumping into someone you know at the supermarket. You can’t expect it to happen each time you go, but over time it’s likely to happen sometimes.
This is why instead of being told about your revenue per impression by ad platforms, you are generally told about your Impression RPM – as in your revenue per one thousand impressions.
[Note: I personally prefer Page RPM as a website publisher because ad impressions are frequently blocked or broken. As far as I am concerned I am outsourcing my advertising to an ad platform, and so making sure that doesn’t happen is their business. By focussing on page RPM instead I am looking at how much I get paid per thousand page views – and that is something I can affect.]
Reasons why you might get less revenue from more ad impressions
So keeping in mind that the value of ad impressions depends on many things, you should still more or less expect more ad impressions to equal more revenue.
However, this is not nearly always the case. Some of the reasons why the (potential) value of ad impressions fall are your fault, and some are not.
The main overarching reason however is competition. The more advertisers that want to advertise on your site, the more bids there are in an ad auction. And more competition = higher bids.
Therefore here is a list of reasons why competition for ad space on your site may drop:
- Performance on your site drops. This could be to do with something you have done (like breaking your site or adding too many ads). Or something your ad network has done (like breaking their algorithm or your ads). If advertisers can’t get what they need from your site at a good price, they will move on.
- Your site has content on an ‘undesirable’ subject. If you publish an article on any subject that advertisers might not want to appear besides (for example religion, politics, adult content, guns, accidents) then your site might get tagged as inappropriate and fewer advertisers will book ad space on your site.
- There are too many ads on your site, or they are too intrusive and annoying. Many advertisers have signed up to an agreement such as Acceptable Ads or Better Ads, where they will only advertise on sites which aren’t overloaded with ads.
- Your ad network/platform is having a bad month. This can happen for no reason you are aware of – maybe a member of staff leaves that was doing all the work or something like that. Sometimes there are just bad months.
- It’s the middle of the month (or quarter). Big advertisers plan their advertising budget in advance, so at the start of every month lots of ad campaigns start at the same time and this drives up prices. Similarly, at the end of the quarter, lots of advertisers will have leftover budget, so there is usually a mad scramble to spend it and get some more results before they have to write their quarterly reports.
- Something changed with ad blockers. Perhaps a celebrity mentioned ad blockers and loads more people installed them. Perhaps a large ad blocker improved its technology and is not blocking more ads. Ad platforms and ad blockers are locked in a forever war and it’s hard to know what is happening between them at any time.
- There is a recession. Marketing budgets are often the first to get cut in a recession, which means fewer advertisers are competing for ad space, so the price drops. If the recession happens in a country where you don’t live, but from which your site gets lots of views you might not be aware of it at all.
- There is a new ad platform. For example, TikTok is really eating into the amount of money spent on Facebook Ads. If there is a shiny new way to advertise then advertisers might flock there instead of coming to your site.
- One (or many) of your largest advertisers disappear. If you have a core set of advertisers earning you money, something can happen to change the status quo. Perhaps they go bankrupt for some reason. Perhaps they stop working with your ad network. Perhaps they all find somewhere better to advertise.
- There are new laws or regulations that change how the industry works. Again, if these happen in another country, you might not be aware of them but they could still impact you.
- It’s sunny. This one is ridiculous but true – when it’s sunny people go outside more so there are fewer people on the internet. As this will reduce your ad impressions, it will also lower your chances of giving an advertiser what it wants.
- It’s a national holiday somewhere (which has the same effect as when it is sunny).
- Random world events. Sometimes undetectable trends emerge for no understandable reason. Maybe someone made a great joke about clicking ads so trust in ads went up. Maybe a new design of ad became fashionable and so clicks across the board went up. Sometimes things just happen. Everything fluctuates – look at monthly stats rather than daily (and compare them Year-on-Year) and you will get a truer picture of what is happening.
“
[This was originally on Quora]
See you next month,
Justin
Enjoyed this edition of OAA? Get it & great articles from industry experts straight to your inbox with our monthly roundup email >