Online Advertising Answers: April 2020

Q: “What are the worst ads?“

&

A: “Hi Anish

In my opinion, the worst ads online are ones which are:

• Dishonest (say they are for one thing but are for another)
• Broken (or break the page you are on)
• Annoying (auto-play videos or sound, pop-ups, mid-roll videos etc)
• Use bad functionality to force a click (such as small close buttons)
• Are straight up tricks to get you to download a virus or fall for a scam

These types of ads are ruining the internet not only by making it a worse place to use but also by making advertising a less viable method of earning money for websites.

As I’ve said elsewhere on Quora – advertising revenue makes so much of the internet free to use. By making ads that everyone hates (more than usual), the people who run these ads are pushing the internet to be less diverse and more expensive.

A Terrible Feedback Loop

You would assume that annoying and tricking people would eventually drive away enough customers that people would stop running these terrible ad types.

Unfortunately, that is not the case. All of these types of terrible ad will have artificially good metrics.

For example, if you go to a legitimate site to download some real software, there may be an ad which is just a download button on a blank background. Clicking it will either download some malware or take you to a dodgy site. As much as most people learn to ignore these sorts of tricks, they work enough of the time that they persist.

Another good example is that if a video auto-plays, then you might click to pause it and accidentally click through to the advertiser’s site. Or if there is a teeny-tiny close button on an overlay you might be driven insane trying to get it to work (and also accidentally go to the advertiser’s website).

And because these bad ads get a reasonably good amount of clicks, the algorithms for the ad platforms they are on will think they are good and popular ads – and show them more often (yes – it works just like Facebook most of the time).

Ad Blockers Are Not The Solution

The simplest thing you can do to stop bad ads is to stop seeing ads altogether of course. There are many browsers which do this automatically, and there are many plugins you can get which do this for the browsers which do not.

Using ad blockers is a temporary (and selfish) fix, however.

To start with I would like to mention that adblockers are basically running a shakedown on websites. Their business model is not generally to be good at blocking ads, it’s to charge websites and ad networks to be allowed to show ads.

Ignoring that problem, by blocking ads on websites you are reducing their revenue. If you are blocking ads on a site you don’t like or are only visiting the once then perhaps that doesn’t bother you. If you are blocking ads on sites you do like then you are really cutting off your nose to spite your face.

This is because if a site loses ad revenue, they generally don’t take that as a sign to move away from advertising revenue. They will instead try short term fixes to mitigate the issue – such as adding more ads to their pages or allowing more annoying ad types.

But “aha I block ads so this doesn’t affect me” I hear you cry – well in the short term sure. It will drive away other users who haven’t blocked ads, however, and this will hurt the site even more.

Eventually, a site with lots of users which is struggling for cash will probably do one or more of the following:

• figure out how to get around adblockers (so now you will see all the old ads PLUS the extra annoying ads they allowed in the meantime)
• start charging for content and put up a paywall
• run ads that can’t be detected (such as native ads or product placement – and both of these are MUCH more annoying than display ads)
• cut costs (perhaps by firing your favourite writers)
• sell to another site
• go out of business

So as I said – using an ad blocker on all sites including ones you like is really just making the internet worse in the end.

And let’s not forget – you can’t block ads on Facebook, so by blocking ads on your favourite site you are just driving more ad revenue to them.

And let’s be honest – you didn’t even actually notice most ads to start with, did you? Most people don’t anyway! So what are you even blocking really?

So what is the answer?

COMPLAIN! For all the is good and great about the internet, please complain whenever you see a bad ad. Get in touch with the site running it – take a screenshot if you can and send it to them, and tell them which country you are in, and perhaps your browser and the time of day and say there was an ad you hate running (and why).

They will take you seriously. There are essentially infinite ads out there for sites to run, and so SITES DO NOT MIND BLOCKING INDIVIDUAL ADS. They want to make money out of running ads *and* they want you to love their site. So if you see an annoying ad, tell them and they’ll almost definitely get rid of it for you.

People on the internet complain about everything, but rarely about ads for some reason. The more complaints there are, the more websites and ad networks will take a look at the ads they run, and the less bad ads they will run.

Also, and more importantly, if you are a website owner you should do whatever you can to not run annoying ads. Don’t be tempted by the short-run boost in income that dodgy ads will bring. Always prioritise your users, and you’ll make more revenue than you otherwise would.

Sorry for the rant, but I hope this helps somehow,

Justin“

[This originally appeared on Quora]

Q: “How much sales can we get through Google Ads?“

&

A: “Hi Ajay,

If I may, I would like to answer your question with a song…

Depending on how much you spend, and how good you are at advertising, there is effectively no upper limit to how many sales you could get through Google Ads (especially if you are selling worldwide).

I mean – I’m sure there are technical limits (like you could sell so many that there would be no money left in the world to buy more, or everyone in the world could own as many of your items as they can physically store in their house, or you could use up all of the world’s resources so you can’t manufacture any more), but these obviously aren’t real issues you need to worry about.

The real question is, what is the maximum amount of sales you can get from Google Ads which are still earning you money. To work this out you should consider your ROAS (and your ROI).

Conversion Value / Cost (aka ROAS – Return On Ad Spend)

If you want to see if you are earning more from your Google Ads than what you are spending, then you can simply work out your ads direct profit.

To do this you simply subtract the amount you have spent on Google Ads from the amount you have earned from your Google Ads. If the number you get back is above 0 then congratulations – your ads are profitable!

*****

Note: This assumes that you have set up eCommerce tracking in Google Ads – which is a must if you are trying to get sales. The best way to do this is to import goals and transactions from Google Analytics. See Google’s guide on how to do this here.

*****

Unfortunately, Google Ads does not have a “Profit” column available in its reports. What it does have is “Conv. value / cost”. What this means is the conversion value (ie how much you have earned from your sales) divided by the amount you have spent on Google Ads.

What this is actually showing you is how much you will get back per dollar (or Pound, Euro, Rupee, etc) you spend. So for example, if your Conv. value / cost is 2.5 that means you are earning $2.50 for every $1 you have spent on ads.

To put it plainly – if the number is above 1 then you are earning a profit.

While Google Ads refer to this metric as “Conv. Value / Cost” it’s actually a standard measure for advertising (you can find it in Facebook Ad Manager for example), and is usually called ROAS – which stands for Return On Ad Spend.

Note: ROAS is expressed as a percentage, while Conv. Value / Cost is a number. To switch between the two, divide ROAS by 100.

Click to enlarge

ROAS = Amount Gained From Ads ÷ Amount Spent On Ads

[Use this free ROAS calculator to work your Return On Ad Spend]

ROAS or “Conv. Value / Cost” is a more useful metric than just profit, as it answers the question “how much more profit would I make if I spent more on ads?”. If your ROAS is 5, then spending $100 more should get you $500 back (which would be $400 direct profit).

This helps with planning ad campaigns but also helps you to decide where to focus your budget. Move your budget to campaigns / ad groups / ads with the highest ROAS and you will earn more money back from your ads.

Simple! Well not quite…

A warning – results don’t last forever

It should be noted that unlike the song I embedded above – there is actually usually a limit. The more you spend on any individual campaign or ad, the more likely the ROAS is to drop.

This is because Google Ads will find the people most likely to buy from you as soon as possible, so the first lot of budget you spend will likely be the most efficient. If you add more money to the budget, then you’ll be reaching more people who are slightly less likely to convert. The more you spend, the less valuable the audience you will be reaching.

On top of this, the more you spend the more you’ll be showing your ads or campaign to the same people who have already seen it – including people who have already made a purchase.

Together this means that adding more budget, or running a campaign for longer periods generally means a decrease in ROAS over time. A high ROAS is still a good sign and still shows you where you should invest your advertising budget. However, you should always be trying different messages and different audiences to keep your marketing robust.

Sometimes you simply have to pause a campaign for a few months in order to give your audience a break before bringing it back. Sometimes your messaging just stops resonating.

Therefore you should use ROAS as a dynamic metric – always have multiple campaigns, ad groups and ads running, and keep testing to see if you can improve on your performance over the past 30 days. If something falls behind, switch it off and try something new.

Sounds good, but how do I see my ROAS in Google Ads?

To add ROAS (or Conv. Value / Cost as it’s called) to your Google Ads interface, follow these steps:

1. Click on the columns button above your results table in Google Ads

2. Click modify columns from the drop-down menu that appears:

3. Click on “Conversions” and in that section, tick “Conv. Value / Cost”.

*****

Note: There is a section for Conversions here and another section for All Conv. This is because you can choose which conversions appear in your conversions column. If you want to understand why and how to do this – read Google’s guide here. If you haven’t done this – then both the tick boxes mean the same thing.

*****

Please keep in mind that when you add columns to Google Ads, they appear on the far right of the set of columns you already have showing, so you may need to scroll right to see them.

When does ROAS not help?

ROAS can tell you which ads are the *most* profitable but doesn’t strictly tell you if those ads are bringing in a profit.

ROAS only takes into account the amount of money you have spent directly in Google Ads (or whatever ad platform). However, ads cost more than just that. They take time and money to make, build, and run (and sometimes to host and track too). On top of that, there are all the costs of whatever it is you are selling.

So to calculate if your ads are actually making you a profit, you should use ROI instead of ROAS. ROI (Return On Investment) takes into account all your expenditures, so you can see how much something actually earns. ROI, unfortunately, does not appear in Google Ads at all (because it takes into account things which happen outside of Google Ads).

Click to enlarge

ROI = (Amount Gained – Amount Spent) ÷ Amount Spent

[To calculate ROI, you can use this free calculator.]

ROAS also does not help you understand the long term value of acquiring new customers – at least not without some further calculations.

For example…

Sometimes sales are used as a “loss leader” meaning that you sell something for less than it cost you to make in order to get customers attention (or simply into the habit of buying from you). This can be very effective if people are wary of purchasing from you and you just need a chance to prove yourself. It can be a costly strategy (of course) but in this scenario, ROAS wouldn’t be helpful – conversion rate would be a much better metric as you just want as many sales as possible.

ROAS also does not take into account the branding power of sales. Once someone has purchased from you, they have crossed the invisible barrier into trusting you and so are more likely to buy from you again. So a single sale is often worth more than just the profit from that sale – it brings with it a customer and whatever their lifetime value is. For most businesses, this LTV metric is very difficult to calculate, unfortunately.

A bonus warning – there are usually limits

Your question didn’t specify anything about your business so I have been talking hypothetically about using Google Ads to sell.

However, for most business, there are natural limits to how much you can sell on Google Ads, such as:

• Not everyone wants to buy your product (I don’t think any products are universal)
• You can’t sell your product outside of a specific geographic area (due to laws, or issues with shipping etc)
• Your branding / marketing doesn’t work on everyone (I don’t think any marketing is universally effective)
• You most likely have competition, and they will get some of your potential sales
• Marketing budgets aren’t infinite, even when they bring in more money than they spend
• Google Ads can be tricky, and your success depends in a large part on how skilful the person who is running it for you is

And of course, the endless, infinity variety of other factors that affect running a business.

In conclusion

You can (theoretically) get as many sales as you can afford from Google Ads.

To make this viable you need to set up conversions and import transaction values from Google Analytics.

To make this efficient you should optimise towards ROAS (called Conv. Value / Cost in Google Ads) to make sure your ads are earning you as much as possible. A score under 1 means you are losing money.

To ensure you are actually profitable, you should keep your eye on your ROI.

BUT if you only want sales no matter if they lose you money in the short run, you should optimise towards conversion rate instead of ROAS

I hope this helps,

Justin“

[This originally appeared on Quora]

Q: “Why is my CTR amazing, but my CPM is awful with Facebook ads?“

&

A: “Hi Luke,

You are probably getting an amazing CTR and awful CPM on Facebook because a higher CPM means you are paying to access a ‘better’ audience.

It’s very likely that this audience is more expensive because there is more competition for it, and that there is more competition for it because it drives higher CTRs.

A low CPM probably means a low CTR

I’ll put it this way – Facebook doesn’t want to say no to your money, no matter how much you spend. If you pay a 1 penny CPM, they would ideally still want to take that money from you because they have so much advertising space available.

Facebook isn’t going to give you a good audience for 1 penny CPM though – at that price they’ll send you the audience that no-one else wants.

And it’s likely no-one else wants that cheap audience because they are very unlikely to click on ads.

Facebook saves access to audiences who are likely to click for the advertisers who pay more.

This is not only because they want to encourage those advertisers to come back, but also because over time the best audiences have been identified and are fought over by advertisers. This drives up the price.

It’s like with any inventory being sold – scarcity increases price. In this case, the thing that is scarce is good audiences on Facebook.

CPC or CPA are better measures than CPM

You shouldn’t really be too concerned with your CPM on Facebook anyway, to be honest. The number of people your ad reaches is a secondary concern to what your ad actually does.

If you are selling something or trying to get people to take an action, you should really be working on a CPA basis. This means seeing how much your advertising costs you per action (CPA). To do this you’ll need to set up a Facebook pixel (unless you’re just running lead gen ads), which Facebook explains here: Help Center

[Learn more about CPA advertising here.]

If you can’t add a pixel to your site, or you are simply looking for clicks, then you should really be looking at your Cost Per Click (CPC). This way you can still judge your campaign entirely on results.

A lower Cost Per Click definitely means more clicks for your money, however again, Facebook will be sending the best clicks to people who pay more. There is a way around this though…

How to improve your CTR and CPC at the same time

There is an easy way to identify audiences that perform well for less money. It works on anything really – ads, age targeting, or whatever you split out.

It’s a simple metric I made up which I call VfM (Value for Money).

Click to enlarge

VfM = CTR ÷ CPC

Simply divide your CTR by your CPC. Or else you can use this VfM calculator to save time and experiment with different scenarios: Value for Money Calculator

How to use VfM to improve performance

Simply split out audiences into different Ad Sets, and then use this metric to compare them. What the VfM value literally means is the CTR you are buying with your CPC.

To put it another way, VfM means what CTR you get per $ you spend on a click.

The way it works is simple – if you increase the CTR then the VfM number goes up. If you reduce the CPC then the VfM number goes up.

It’s not really about the maths – more just about what changing the parts of the equation mean. Positive results for either your CPC or CPM make the VfM number higher!

You should really be using CPA advertising if you can. If you can’t, this VfM formula will help you get more out of your advertising when you pay for clicks.

I hope all this helps.

Justin“

[This originally appeared on Quora]

See you next month,

Justin