Advertising Payment Types
What do you choose? CPA vs CPC vs CPM vs CPE? PPA vs PPC vs PPM vs PPE? What are they and what’s the difference between each? Is CPM better than CPC? Does CPA make more money than CPC? As a publisher, what’s better when comparing CPM to CPC? What about as an advertiser? Here we explain all that and more.
Pay Per Action (PPA)
Under the PPC model, a publisher is paid when an eyeball completes an action after clicking on an ad. Each action is therefore worth $X per action, often labelled as $X CPA for “cost per action”.
Pay Per Click (PPC)
Under the PPC model, a publisher is paid when an eyeball clicks on an ad. Each click is therefore worth $X per click, often labelled as $X CPC for “cost per click”.
Pay Per View (PPM)
Under the PPM model, a publisher is paid when an eyeball views an ad. PPM stands for Pay Per Mille, where Mille is French for a thousand. This is typically measured in thousands of views. Each thousand views is therefore worth $X per thousand views, often labelled as $X CPM for “cost per mille”. The reason for the thousand view unit is because the price per page view can often be under $0.01, so looking at prices per thousand gives a figure people may be more used to. For example, $3 CPM compared to $0.003 per view.
Pay Per Engagement (PPE)
Under the PPC model, a publisher is paid when an eyeball engages with an ad. This may mean hovering their mouse over an ad (where then an expanded ad may appear) or some other criteria specified in the agreement. Each engagement is therefore worth $X per engagement, often labelled as $X CPE for “cost per engagement”.
PPA vs PPC vs PPM vs PPE
Also known as the CPA vs CPC vs CPM vs CPE debate
Factors to Consider
The three main factors to consider are price, risk, and transparency.
At an attractive enough price, any of the methods could be attractive.
For example, if the creative was horrible and the click-through-rate (CTR) was super low, getting paid per click as a publisher could still be worth it if every click was being paid out at $2 million. This is an extreme example, but you get the idea.
The different methods vary in risk for each party.
Pay Per View is least risky for a publisher and most risky for advertisers.
The publisher gets paid for every ad impression, regardless of whether it results in a sale, click, install, or even if the eyeball looked at the ad consciously (may have just scrolled past). As long as the ad is displayed, the publisher gets paid. Sometimes this depends on the eyeball’s browser actually getting to the adspace, but the argument still holds.
The advertisers pay for every ad impression regardless of whether the creative performs. Therefore it puts more onus on the advertiser to have good creatives and ensure the traffic quality is good in order for it to convert.
Pay Per Click is similar to the middle ground in terms of risk.
The publisher gets paid every time an eyeball clicks on an ad. If a creative from the advertiser is not performing well, the publisher may be serving a lot of ad views but not getting paid for them.
From an advertiser’s perspective, this filters what they’re paying for to users who are supposedly interested in the advertisement. This creates an incentive for advertisers to design creative so that only users who are very interested will click on an ad.
Pay Per Sale is least risky to advertisers and most risky for publishers.
The advertisers only pay when a sale is made. Assuming that sale results in net profit after taking into account all costs including the amount paid to the publisher as commission, the advertiser comes out ahead. For example, if the advertiser is selling a non-refundable $20 flashlight, costs excluding advertising is $5, and it pays the publisher $5 per sale, the advertiser’s position looks very de-risked. The publisher therefore is assuming most of the risk. The publisher can serve a lot of page views but the publisher won’t be paid if the eyeballs don’t complete a sale. The publisher can send a lot of clicks but the publisher won’t be paid if the eyeballs don’t complete a sale. In a fraud situation, the publisher can send a lot of buyers but still won’t be paid if the advertiser doesn’t report all the sales.
The amount of transparency is also a sliding scale for advertisers and publishers between the multiple methods.
The publisher can track views and clicks. The publisher cannot track if a sale is subsequently made and for how much, if an app is installed, etc. Therefore the publisher only has visibility on the number of ad views it has served and the number of times those ads have been clicked on. The publisher relies on the advertiser for information on whether an action has been completed (sale/install/lead), and there could be fraud from the advertiser side on under-reporting these actions.
The advertiser has visibility on impressions, clicks, and actions. Actions are completed on its own website/service so naturally an advertiser can track this. For example, if the advertiser is a shoe store, and a customer places an order after clicking on an ad, the advertiser knows a sale was made in addition to what was purchased and for how much. The advertiser also has visibility on clicks in that every (valid) click made results in traffic to the advertiser’s website or service which can then track it. Typically it goes through a tracking URL that scrapes information about the click as well. Depending on whether the publisher has outbound tracking URL, the information gleaned from the click may be partial. The advertiser may also have partial transparency on views if the creatives or tracking code is to their own servers. By partial transparency, it means that the advertiser knows when its tracking service has been “pinged”. It does not know whether the eyeball actually looked at the ad or if the ad was actually served properly. For example, if the publisher set up code to ping the service but hide the ad to users who are logged in, the advertiser has no way of knowing this.
Example of CPC vs CPM
Suppose an advertiser is selling lawn mowers and compare these two advertisements. The first one says “You won’t believe what this lawn mower does” and the second one says “Want to buy a $399 German lawn mower?”. The first one will result in a lot of clicks but not a lot of conversions into a sale. The second one should theorectically result in lower clicks but of those clicks much higher conversions into sales.
Suppose the advertiser pays for 1000 views, and the first creative generates 200 clicks and the second creative generates 20 clicks.
If an advertiser is paying per view, the advertiser would likely prefer the first creative. The first creative would lead to a lot more clicks and traffic to the advertiser’s website than the second creative but the advertiser is not paying extra for those additional clicks. The amount paid is fixed regardless of how many people click through, so the first creative leads to ten times more clicks (and traffic) for the same price.
Suppose 1% of the people who click on the first creative buys a lawn mower and 5% of people who click on the second creative buys a lawn mower.
If an advertiser is paying per click, the advertiser would likely prefer the second creative. For the same price per click, users are 5X more likely to buy a lawn mower. This means the advertising cost per lawn mower sold is 80% lower for the second creative compared to the first creative when paying per click.
If an advertiser is paying per view, we’ll have to do some math. The first creative generates 10X more clicks but 5X less conversions per click. This means that the first creative would be chosen as it leads to 2X more sales for the same number of ad views. This choice would change depending on the click through rate (CTR) and the conversion rate per click.
As you can see, pricing pays a large part in what method to choose.