Price floors seem like a great concept for publishers. They choose a minimum amount they are willing to be paid to display an ad and if there are not any available ads at or above that price then they do not show one. In theory this makes a lot of sense and will protect an advertiser from serving ads at a very low price and thus not making much money. This approach works well for publishers with premium properties because they can demand higher CPMs (price per 1,000 ads displayed) but for newer and non-premium properties it can have a detrimental affect.
The downside to floor prices is that it can adversely affect the fill rate if the floor is not set properly. If the floor is too high, advertisers may decide to not fill the ad request and the publisher could miss out on potential revenue. The alternative option is to do a rev share. This could potentially result in a lower CPM but a higher fill rate and thus more revenue, which ultimately should be the publishers main goal.
Here is a simple example to show how a rev share agreement can benefit publishers:
In scenario 1, a publisher has 10,000 video ad opportunities with a price floor of $5. At a $5 price floor, 3,000 of the 10,000 ad opportunities are filled for total revenue of $15 – (3,000/1,000)*5.
In scenario 2, the publisher still has 10,000 video ad opportunities available, but has a 50/50 rev share. With this rev share setup the publisher’s ad slots are available to fill at a lower price point so the fill rate is much higher. In this scenario, 8,000 of the ad opportunities fill at an average CPM of $4 (some of the requests fill at a higher CPM and some at a lower one but they average out to $4). The total revenue in this setup is $32 and with a 50/50 rev share the publisher ends up with $16, a dollar more than with the price floor even though the CPM is lower. A dollar more may not seem like much of an increase, but when accumulated over thousands of impressions it can add up.
There is no guarantee that a rev share arrangement will fill better or will increase a publisher’s revenue, but a publisher should be willing to try it if they are not happy with the results from using a price floor. The results can vary widely depending on the publisher’s industry and time of year, so if a publisher does decide to test a rev share, they need to make sure they run it long enough to be able to accurately decide if it is helpful or not. Many newer publishers go straight to the price floor model without considering a rev share one and that is ok, but more often than not, publishers will be pleasantly surprised with the improved results from using a rev share pricing model.