Online Advertising Answers: June 2020

Q: “How much ad revenue do you make per ad?“

&

A: “Hi Peter,

There are two ways to take this – how much can you earn from selling ad space, and how much can you earn from running ads. I’ll deal with both in turn…

If you mean how much money can you make from putting a single ad on your website, then the average is probably about a third of a penny (USD), and the absolute maximum is probably about 50p (this time in GBP).

The Average Amount An Online Ad Makes

For Google AdSense (which is the most popular display ad network), the overall average Page RPM is around $10. This means that for every 1,000 page views a website has, it earns $10 – or 1000 US cents, so 1 penny per page view.

As most websites have about 3 ads per webpage, that penny is split three ways, so a third of a penny is earned by each ad (on average).

This is, of course, a huge simplification, as many sites have far more than three ads per page, different niches and countries pay different amounts, ad rates are seasonal, and each ad size does not earn the same amount. On top of this (as we are currently seeing) – global events can dramatically change the average Page RPM that websites can earn.

However, on average the amount you should expect to earn from a single ad shown once on a webpage is about a third of one penny.

[To work out how much your website can earn, you can use this free calculator]

The Most You Can Make From A Single Online Ad On Your Website

News organisations generally charge the most for display advertising space, and the most I have seen on a rate card is £90 CPM – which means £90 for 1,000 ad impressions, or 9p per individual ad shown (in GBP).

Most companies keep their Rate Cards to themselves, however, so this is likely not nearly the top rate.

Or course you can also sell sponsorships, which vastly increases the value of an individual ad impression. By selling all the space on your homepage for example you can earn up to 5x the normal amount you would sell it for – making the top amount you could make from an individual ad closer to 50p.

If you mean how much money can you earn back through advertising your products from a single ad, then there isn’t any upper limit as it depends on what you are advertising, and whether you want to consider the lifetime value of a customer.

For example, if you were selling superyachts, and an ad led to a someone who not only bought a superyacht, but had it serviced with you for their whole lives, bought more superyachts now and then, and convinced their friends to buy superyachts from you and so on then, one ad could literally make you billions.

However, if you mean how much are you likely to make back from an individual online ad, then the answer is a bit simpler – and it’s between 2x–25x the amount of money you spent.

The Average Return On Ad Spend

Return On Ad Spend (or ROAS) is how you calculate how much you make from an ad per dollar spent. It is calculated like this:

Click to enlarge

ROAS = Amount Gained ÷ Amount Spent (multiplied by 100 to make it into a percentage)

[Use this free calculator to work out your ROAS]

In general, a common target for ROAS is 400% – meaning that for every $1 spent, you get $4 back. This is across the board though, so focussing on individual online channels gets you quite different results.

For example, according to a study by Nielsen in 2016, the overall average was 287%, but the average ROAS for Display Ads was 263% and for video ads, it was 153% – so you should expect to earn $2.63 from a Display Ad, and $1.53 from a Video Ad.

However, for Social Media and Paid Search things can go a lot higher depending on what you are selling and to who. I’ve often seen a ROAS on Facebook Ads in the 1000–1500% range (although 500–1000% is a lot more common), and from Google Ads, it’s frequently more like 2000–2500%. This means a dollar on Facebook ads often earns between $5-$15 back, and on Google Ads, you generally see more like $20–$25 back for every dollar.

So it completely depends on what channel you are using, and what you are advertising. $1 of ad spend can theoretically bring you an unlimited return, but realistically you should hope for at least $2.87 back, and aim for at the very least $4 back.

I hope this helps,

Justin“

[This originally appeared on Quora]

Q: “How do I estimate ad impressions available for now?“

&

A: “Hi Sean,

The reasons why people click on online ads are a lot simpler than anyone would like to admit.

To start with, the quality of an ad, of course, matters – marketing exists for a reason and there are good ads and bad ads. What makes a good ad is somewhat indefinable though so I’m not going to try and get into it here. What I can tell you is the general mechanics of what makes people more likely to click on an ad.

Bigger = Better

The most depressing secret about online advertising is how much size matters. The more space you take up on someone’s screen, the more likely they are to notice your ad. This is why the 468×60 banner ad was replaced with the 728×90 leaderboard ad at the top of most people’s websites – simply because it is larger and therefore works better.

It’s also why the 120×600 skyscraper unit was replaced by the 160×600 wide skyscraper ad unit, which was then replaced by the 300×600 Half Page Unit. Taking up more space on webpages earns more clicks. Newspapers charge more for larger ads for the same reason.

This goes double in search. In Google Ads, you can attach all sorts of “Extensions” to your ad. Ad Extensions are little extra bits that you tack onto your actual ad (such as your phone number, your address, your slogan, etc). Each of these are additions to your original ad, and the more you have running with your ad, the more space in a search result that you get.

The bits in the red boxes are all ad extensions (and possibly the end part of the end in each ad).

If you want proof these make a difference – the largest of the ad extensions is called a sitelink extension, which is a sort of mini-extra ad leading deeper into your site. You can add up to four of these to your ads at a time and according to Google, they boost CTR for an ad by 10–20% (or up to 50% on a branded search term). That is an insane boost in clicks.

Want more proof? If you run a Google Ad on a search that you already rank #1 for organically, you will get more clicks than if you added together with the results of the organic result appearing alone with the results of your ad appearing alone. It’s a sort of 1+1=3 effect. It doesn’t exactly make sense, but it’s true nonetheless. Size matters for online advertising.

People Have To (Definitely) See Your Ad

This one seems obvious, but it’s weirdly not when it comes to online ads. We usually measure impressions online to say how many times an ad/result/post has been loaded. But being loaded onto a website is not the same as being seen by a person.

Think of it this way – when you see (broadsheet) newspapers stacked on a shelf, you can only see that top half of the front page (because they are folded in half). This means that whether or not you buy that newspaper, you can see everything that is on the front page which is above the fold.

This makes that space the most valuable advertising space a newspaper has to offer – simply because more people see it. Similarly, anything below the fold in the paper is only seen by people who keep reading past that initial burst of content.

The same is true on the internet. The “fold” for websites is the bottom of a computer screen when it first loads. As computer screens are all different sizes, the general rule is to get people to see your ads is to place them as high as possible on webpages.

If people actually see your ads then, of course, that means they are infinitely more likely to click on them – so above the fold ads get the most clicks.

[FYI Ad impressions which appear at least 50% onscreen for at least 1 second are called “Viewable Impressions” – buy them instead of regular ad impressions whenever you can as they perform much better]

Seeing Ads Enough Times (But Not Too Many Times In One Place)

Traditional marketing wisdom states that you need to see an ad about 5 times before it really starts to work on you. These are called Opportunities To See, and were the basis of offline marketing for years.

However, at the same time, the most effective frequency cap to improve performance for display ads is 3/24 – meaning that a single user only sees an ad at most 3 times over 24 hours.

These two things together mean that you ideally want people to see the same marketing message in multiple places for it to really take effect. For example, you could use display ads to create awareness, social ads to engage, and email or search ads to convert.

By getting people to see your ads over and over again in different places, you are making it so people are much more likely to click. Think of this way – the first time you hear about something you might dismiss it out of hand. If you keep hearing it again and again from different sources, you might take it as a given.

It’s like a form of pester-power but for advertisers.

via GIPHY

To put it another way – think about engagement rate, and why it’s such a big deal as a metric. An engagement is simply any interaction with an ad/post. In itself it seems pretty meaningless – someone clicking anywhere (including to pause videos, or just see the rest of a chunk of text) doesn’t tell you much. However, what it does tell you is that person has noticed your ad/post.

And that is a big deal for marketers. Getting people to notice your message 5 times means that a person is much more likely to buy what your selling

Everything In The Right Place

When online advertising started, all there was for targeting was context. You’re advertising bikes? Cool, let’s put it on this bike website. No space left on bike websites? Cool, this other website about running mentions bikes a lot too, so let’s advertise on this running website instead.

And it worked. It still works now.

The reason advertising as a whole mostly switched to programmatic/algorithmic/behavioural targeting is not because it works better, it’s just that it requires less manpower. Machines can easily look at quantitative data (such as person x visited this website and is browsing from this country, so they probably like this product) and work out what to sell. What machines can’t do (very well) is see a sites qualitative qualities (eg a site being beautiful, or highbrow, or fun) and work out what to sell on it – that takes people.

And understanding indirect context is what makes a lot of advertising successful. for example, people like having a cup of tea while watching soap operas, meaning that websites about soap operas are great places to advertise tea. Seeing an ad and feeling like it makes sense lowers your defences and makes you more likely to click it.

Sometimes context can be created by advertisers (think eating popcorn in a cinema), and sometimes it happens naturally (think eating burgers + watching sports). Either way, essentially some things just feel right together.

Nudging You To Buy Something You Were Already Thinking About

You know when you go to a shoe website (for example), put some shoes in your basket but then don’t make a purchase – and subsequently get followed around the internet by ads for shoes? That happens because you have just designated yourself in the market for shoes, so shoe seller look at you as a soft target.

And to be fair they are right. What’s more likely – someone who has never looked at shoes online suddenly buying some shoes, or someone who was literally just looking at shoes buying some?

I personally hate this type of advertising, not least because it’s creepy and makes people hate advertising even more than they do already. However, it does work.

Just… data

The most likely reason for you to click an ad is that marketers (or algorithms) have worked out somehow that people like you are likely to click that ad. This optimising of ad campaigns essentially involves looking at data and spotting trends.

More clicks on a Tuesday? Cool, show more ads on Tuesdays. More 34-year-old French vegans who live in Spain clicking at 2am on a Thursday – sure show more ads to them.

The more an ad runs, and the more data that is made available from an ad platform, the more that advertisers can find people similar to you, and show you ads that they clicked on.

We all like to think we are unique individuals, but if I run a set of ads and show them to 1,000 people in England who like Football and Star Trek, then the number of people who click on it out of those 1,000 people will be about the same as the next thousand people. Humans react fairly uniformly to ads (when you drill down enough to find common traits).

I’ve been in digital marketing for almost 20 years now and I still find that fact super strange.

Some People Just Like Clicking On Stuff

The stat used to be that 90% of all online ad clicks came from 10% of people, although it now differs from channel to channel. The people doing all this clicking do make a lot of online purchases so their clicks aren’t worthless – but they also click so much that they skew CTRs across the internet. If you are buying cheap clicks for an ad campaign – it is very likely that they will be the ones doing the clicking.

It’s nothing nefarious – they just don’t worry about where their mouse goes. the demographics I have seen say they are also into online gambling, and play a lot more online games in general than anyone else. They just like clicking on stuff ok!

While I don’t personally judge them for living a worry-free internet life, it is annoying how these people have been somewhat weaponised by online advertising companies. A friend of mine recently spoke to a Facebook advisor who told him to advertise to 18–35-year-old women to get his CTR up (even though they are not his target market) because “they will click on anything”. This was a Facebook adviser telling him how to game the Facebook algorithm by buying up a ton of worthless clicks. For shame!

In Conclusion

Ignoring for the moment how good or bad ads are, people are more likely to click on an ad online for these reasons:

• The bigger the ad, the more clicks it gets.
• Ads which are definitely seen by users are more likely to get clicked (duh).
• Showing people ads more often makes them more likely to click (3 times in one place at most per day, 5 times overall at least).
• Putting ads in places where the ad makes contextual sense makes people more likely to click.
• People in the market for something are more likely to click
• People just like you clicked on something, and so it’s a good bet that you will too
• Some people will just click on anything
• Random happenstance – sometimes a combination of things makes an ad work on you. Life is messy like that.

A final note – more intrusive ads get clicked on more often, but also drive away people from sites. I’m talking about pop-ups, videos ads that auto-play with the sound on, overlays etc. Everyone is more likely to click on these ads, but not for positive reasons, and if everyone could stop using them that would be great thank you.

I hope this helps,

Justin“

[This originally appeared on Quora]

Q: “What does it mean to have a negative relationship between CTR and online purchases?“

&

A: “Hi Zeineb,

Having a negative relationship between CTR and online purchases means that the thing that gets your audience to the sales page isn’t working well with your actual sales pages.

Either your sales pages aren’t very good, or simply the messaging on your ads / posts / whatever isn’t lining up with what it says.

For example, your ad could say “Buy Dogs Now! $1.99” and so loads of canine lovers click your ad, but it turns out you are selling hotdogs.

It’s not good enough to have an ad / post / whatever that gets people clicking, you need to get the right people clicking.

If you’re paying for clicks on ads, then you may not get sales

It’s actually not uncommon to have a negative relationship between CTR and online purchases, especially if you are running CPC (or PPC) ads.

When you are paying for clicks, the place you are advertising (eg AdWords, Facebook, a random website or ad network) will simply work towards getting you as many clicks as possible, while using the least ad impressions.

This means they don’t care about the quality of the clicks they are sending you. This is especially true if you are paying as little as possible per click.

Also – the majority of clicks on pretty much everything on the internet are done by a comparatively small amount of people who will click on anything. If you are paying as little as possible for clicks, it’s likely these are the people who are doing the clicking.

To get around this you could try to pay more per click, or at the very least not pay the absolute minimum for clicks. This may not solve your problem, but not limiting the amount you pay per click will likely improve their quality.

Are you tracking conversions?

If you are working towards sales, you shouldn’t really be focussed on clicks or CTR anyway. Have you set up conversion tracking? If not, then do so immediately, and try and switch to CPA advertising if you can.

CPA advertising means Cost Per Acquisition. It’s when you only pay if the ad makes a sale. It can be a bit tricky to set up comparatively but is very much worth doing.

Check the options wherever you are advertising or talk to a sales rep about setting this up.

Summary

To improve the relationship between CTR and online purchases you should:

1. Work on your sales page and make sure the message you put in your ad reflects what it says. People need consistency.
2. If you are paying for clicks, consider paying a little more per click. Don’t set a low maximum CPC as it actually won’t help.
3. Make sure you are tracking conversions and letting whoever you are advertising with track them too.
   1. If you are using AdWords, then create goals in Google Analytics, and add that goal into AdWords.
   2. If you are using Facebook or any other advertiser, see about adding a conversion pixel to your site.
4. Switch to CPA advertising if you can – only pay for conversions. If you are running AdWords ads, consider starting on enhanced CPC first.

I hope this helps,

Justin“

[This originally appeared on Quora]

Keep the questions coming,

Justin