You know I don’t tend to get dramatic in my blog posts. The title of the post isn’t meant to sound like clickbait or a desperate scream for your attention. It’s merely stating a simple fact that most advertisers don’t tend to talk about. Which is that the pure existence of pay per click advertising (PPC) is getting closer and closer to complete extinction. If you don’t know what PPC ads are here is a quick explanation:
“PPC stands for pay-per-click, a model of internet marketing in which advertisers pay a fee each time one of their ads is clicked. Essentially, it’s a way of buying visits to your site, rather than attempting to “earn” those visits organically. Search engine advertising is one of the most popular forms of PPC.”
Google Adwords is probably the most recognized model that uses the pay per click model. It wasn’t the first to introduce the model of buying traffic based on cost per click, but when you talk about PPC ads, it’s largely associated with Google Ads.
PPC was introduced before modern tracking information evolved into the form that is used today. Let’s examine the fundamentals of PPC ads.
Fundamentals of PPC
Regardless of the advertising platform, most PPC models use the following structure:
- Select your campaign settings such as target search terms/interests, demographics, creatives, etc.
- Set how much you want to pay for each click and adjust the price according to the performance of your search terms/interests and your ads.
- The price per click is based on a bidding action where other advertisers set the amount they want to pay per click and the highest bidder wins.
Pretty simple right?
But let me ask you a question. Why would you want to bid for a cost per visitor, if getting visitors is not your main goal? Your main goal is sales or leads. Instead of bidding on how much you want to pay for a single visitor, wouldn’t it make more sense to bid on how much you want to pay for a purchase or an actual lead?
Consider this analogy:
Imagine you had to bake a cake and you wanted your oven set at 350° F (180° C). Then, instead of simply selecting the temperature, you have to manually adjust the top and bottom heaters and always measure the heat inside the oven to achieve your desired temperature. In this example, the top and bottom heaters would be your CPC campaigns that drive users to your website and the exact oven temperature is your goal (objective). You constantly need to adjust the heat levels of your heaters to get to the right temperature, right?
Now, consider that, instead of trying to figure out how to manually adjust the heaters to get the right temperature, you simply set your desired temperature and the oven will optimize the heaters to achieve that goal. When it’s too low, it will heat it automatically and when it’s too hot, it will lower the heat.
Wouldn’t it be amazing if you could do the same thing for your advertising campaigns?
Instead of telling the platform how much you want to pay for an action that allows you to reach your main goal (manually adjusting the “heat” settings), wouldn’t it make more sense just to tell the platform how much you want to pay for your real goal and have it optimize towards achieving the goal (the “heat” setting changes automatically)?
Well, that technology has existed for many years and is becoming the primary option to optimize and create campaigns on advertising platforms.
The domination of the CPA model
Cost per action (CPA) is an online advertising marketing strategy that allows an advertiser to pay for a specified action from a prospective customer. It might sound complex but it’s really simple. Instead of setting a bid for a visitor, simply state your desired conversion goal (purchase, lead, etc.).
By telling the advertising platform how much a conversion is worth to you, they can use that information to try to find bidding and delivery opportunities that are most likely to result in you achieving your goal.
The CPA model wasn’t always mainstream. Facebook ads deserve most of the credit for making this advertising strategy the preferred choice for advertisers and, eventually, the other platforms had no choice but to follow.
It’s not like this idea was invented by Facebook. Other platforms like Google offered this bidding strategy, however, for this strategy to be effective, rich tracking is needed to track and learn from the performance. Tracking pixels weren’t as popular on Google Ads as they were on Facebook ads.
When Facebook aggressively pushed their tracking pixels to advertisers, it allowed them to learn a lot about what actions users take off Facebook and use that information to dramatically enhance their targeting capabilities.
By knowing what actions users take on and off Facebook as well as knowing important metrics like your conversion rate, bounce rate, products viewed, and average time on the website, Facebook can predict what users might be interested in. Then, it uses that information to automate bidding and targeting to help advertisers reach their ad goals as well as optimize their ads.
The core advantages of using the CPA model instead of the PPC model.
By letting the algorithm optimize towards your primary goal and not your secondary goal, targeting becomes automated. If, for example, you sell a product that’s for females only and you accidentally select to target both genders and let your campaigns run. Even before you start getting sales, the optimization algorithm will use feedback indicators to see which type of audience is the most relevant for your ads and will shift delivery toward them. This rule applies to all areas of targeting including geographic location and all demographics.
Consider the PPC model targeting scenario. You are telling the platform you want to get the best price for clicks, so you might drive users that cost less but they are also less relevant. For example: If older demographics find your product relevant but have a very low conversion rate and their cost per click is low, they will still get a lot of delivery since you told the platform you want the most clicks for the best price even though their cost per conversion might be high. Think of how many adjustments and maintenance it takes to achieve your results by optimizing for the wrong metrics.
Other metrics become irrelevant
When you optimize towards your main conversion event, metrics like cost per impression, cost per website visitor, or even cost per like become irrelevant.
What difference does it make if a website visitor costs you 10 cents or $2 as long as the cost per conversion is within your target? Cost per impression can go up or down, yet all that matters is your overall cost per conversion.
One common mistake that advertisers make is that they try to lower their cost per conversion by lowering their cost per website visitor or cost per impression. Initially, it sounds rational, because if you need 100 visitors to get 2 purchases on your website and you can lower the cost of those visitors by lowering the cost per website visitor you would save money, right? Wrong.
Trying to manipulate the cost per visitor would simply lower the quality of the visitors who come to your website. Now, you would need more than 100 visitors to get those 2 purchases since the quality is not the same. Cost per click or impression on Facebook increases when the quality of users that you target increases.
The more that your targeting is accurate and fine-tuned, the higher the cost per click or impression might be. Therefore, you will also most likely get a higher click-through rate and conversion rate as well. So, the more targeted your ads, the higher they will convert which means more conversions for you.
How easy it is to manage your campaign budgets and bids when they are all based on your actual goal. When you optimize towards your main goal you can simply set your target cost and a large enough budget for the optimization algorithm to find you as many conversions as possible.
If I set a daily budget of $2000 with a target cost per conversion of $50, I am telling the optimization algorithm that I want conversions for $50 and I am willing to spend $2000 per day if I get conversions for that cost. Then, it will look for opportunities in the market and will try to predict where to deliver my ads and at what price based on my expected conversion rate, click-through rate, and cost per 1000 impressions.
In some cases, the predictions are wrong and my cost per conversion might be higher than my target. However, performance will be adjusted based on the feedback from the previous day and will be adjusted to try to better deliver the ads so that the target goals are met.
To use the CPA model, a lot of data is required. It’s not a model that should be used for a single day of the performance, nor should it be evaluated daily. The CPA model relies on past data performance and works best when you set your target goals and budgets without you interfering daily by making many changes that will require it to constantly update its predictions.
Now, imagine how much time you save versus traditional PPC bidding management. On PPC bidding, you have to constantly monitor the performance of your keywords and adjust their cost per click as some keywords might have a low cost per click but a bad conversion rate, which will result in a high cost per conversion. Again, when you tell the platform you want to optimize by using the PPC model, it will try to find you the best price for clicks, not for conversions.