2020 is almost over, and yet much of our industry still bases calculations of volume, budget threshold and reach on a simple impression basis.
Since 2019, Xaxis Belgium has not been selling campaigns on a CPM (cost-per-thousand impression) basis -- except for one or two products living their final hours. Since it's our heartfelt belief that selling impressions does not mean anything, our display campaigns are sold on a VCPM (viewable CPM) basis and our video campaigns on CPCV (cost per completed view), which provide ways to properly evaluate what impact has been delivered by a campaign.
However, it is still a day-to-day struggle to overcome the reliance on impressions. We are constantly fielding question and comments such as:
But what is an impression without a viewability component? It’s merely an ad rendered in a user's browser thanks to a bit of code. And “rendered” absolutely does not mean "visible.” Viewability is one of the basic fraud checks one should implement in a campaign. The practice of impression stacking is still used by fraudulent publishers and networks.
Even from honest and reliable publishers, purchased impressions are not always "visible.” That's the main reason the viewability rate is so important.
We still have private deals with Belgian publishers that are delivering impressions with a 35% viewability rate, as measured by the MRC standard. This means that among the bought impressions, 65% of them are not made available for at least one second with at least 50% of their display area available for viewing. This is on premium Belgian inventory bought through private deals with established premium publishers. We are not talking about open RTB auctions here.
Without a viewability metric, can we really calculate reach or volume delivered? What is the value of such impressions?
These are the quandaries we are trying to solve by selling impressions on a viewability basis, to give our clients the opportunity to buy only viewable impressions. That way, they will know how many “real” impressions they’ve obtained. They can get an even further assurance by buying according to the GroupM standard, which requires 100% of the ad be in view for at least a second.
Sometimes, using VCPM can deliver a startlingly high price. Have you ever tried to convince a client to buy display viewable impressions at a rate of 17€ VCPM -- a rate that comes up when viewability is as low as in the example given above? You can try using the argument that this is the cost for Belgian premium impressions bought on quality inventory from premium partners. Confronted with such a large number, though, buyers’ reaction is often: "Wow, you're damned expensive!" I’ve even had a few clients say, "But how come Facebook is so much cheaper?!" (That’s a different topic, but let’s for now say that Facebook impressions are not equivalent to those on a premium publisher.)
Buying viewable impressions or video-completed views at a VCPM or a CPCV rate gives a more accurate picture of what value a client has actually received. This will allow them to more precisely evaluate the inventory they are being sold.
Shouldn't we shift entirely to these measures instead of relying on classic impressions? Wouldn't we be better able to evaluate whether we’re reaching precisely the right level of GRPs, get a more precise idea of advertising impact, if a media buy is based on viewable impressions or complete video views?
On top of all the arguments above, we have a marvelous cookieless future ahead of us. This new situation might force us to rely even more on publishers’ reporting.
Will we be able to move further toward the goal of buying inventory based on real outcomes? Or do we step back and keep focusing on impressions?