In our previous article about floor prices, we saw how to set-up a simple yet effective floor pricing strategy to drive maximal ad revenues.
Now that you have put all these recommendations in practice let’s see how you can avoid floor pricing mistakes that can cost you considerable amounts of money.
At the end of this article, we also offer a template that helps you spot issues and discrepancies caused by these mistakes.
P.S. This article only focuses on the implementation in Google Ad Manager.
On the UPR side, when you set-up targeted floors on a part of your inventory (say you are setting up a targeted floor on your 300×250 ad size), the UPR will only apply to AdX and will not apply to Prebid.
This is because Google AdX does not read the right information contained in the Prebid line item regarding the creative. Google Ad Manager reads the size returned by Prebid as 1×1 and not the real size, which is 300×250 in our case.
To solve this problem, you need to create a matching item in your line item, to make AdX read the actual creative size.
To sum up, if you are using size-targeted floors, make sure that these UPRs apply to Prebid too. If you don’t do this, the UPR will apply to Google but not to Prebid, and if you don’t have floors on the SSP side (as we strongly recommend you to do), you will create situations where Prebid will be able to win bids without competition and below Google floors.
Let’s kick-off things with some quick reminders!
When you create a line item set for Prebid, you need to create a line for each CPM value. Since you don’t have infinite time, you cannot create an endless number of line items, and therefore you cannot have an infinite granularity. You will need to make choices, and create an evolutive granularity, which means that you will create (as an example) a line item each 0,10€ until you reach 2€, then you’ll create a line item each 0,5€ until you get to 5€ and so on.
The idea behind this is: the finer your granularity, the fairer the competition between AdX and Prebid. You may ask why?
As you know, AdX participates in an auction with the exact bid while Prebid participates with a reduced approximation of the Prebid winning bid. That is, if Prebid’s winning bid is 0,28€, it will send to the ad server the value of the closest line item under 0,28€ (0,25€, for example), which will be the value that will compete with AdX’s bid.
What we advise you to do to get around this is:
Say that your last line item is set at 10€. This means that when Prebid’s winning bid is above 10€, it almost never wins against AdX, since it participates in the auction with a reduced approximation of the Prebid winning bid, so as soon as AdX bids more than 10€, it wins.
So if you decided to stop creating line items at 50€, then you should be okay, but if you want to stop at 10€ or 20€, you should seriously consider adjusting your line item granularity.The best thing to do to get around this without creating an infinity of line items is to check the percentage of your Prebid winning bids that is above your last line item.
If it is around 2%, then it is useless to create more line items, but if it is higher than 10%, it would be interesting to add more line items. This is something that can easily be calculated thanks to Pubstack 😉
We understand that it can be complicated to go through all the verification process manually. So we are sharing here a template to help you check that your UPRs apply to Prebid, spot issues in the implementation of Google Ad Manager’s UPRs, and spot discrepancies between UPRs and SSP floors.
You can check the template here and download it for your personal use, and please feel free to get in touch or comment if you have any questions about the template.
On top of avoiding the floor-pricing mistakes we’ve just listed, you should also make sure your pricing granularity is properly set up if you want to go the extra mile. If you’re not familiar with them, we’ve got you covered with our article "Unraveling Price granularity and how to retain maximum value".
At the end of the day, superior data granularity & actionability serves one goal : reducing the guess work when it comes to optimising your ad stack. We’ve face this exact challenge with SLM Ads, a publisher whose websites bring together more than 6 million visitors per month. Read our SLM Ads use case to learn more about how Pubstack has optimised SLM Ads’ A/B testing strategy as well as sped up their process towards optimal yield.
Finally if you’d like a deep-dive into how you should A/B test your ad stack, we’ve got just the thing for you : "How to run a proper A/B test ?"