There are plenty out there singing the praises of real-time bidding as a rising tide that will lift all boats. It is, we’re told, the proverbial win-win-win for advertisers, publishers and users. But from my standpoint, I’m not sure that’s the case.
There’s no question that RTB as a mechanism for setting the price for display inventory has its place. Like all auction models, it seems on the surface to offer the best way to guarantee the seller the highest possible price for a product or service, while, mysteriously, also low prices for buyers. But like all auction models – not least because both those things cannot possibly be true at the same time — it doesn’t work that way.
For some inventory, mainly the highly sought-after impressions from users suitable for retargeting, RTB secures a very high price. But for all else, far from raising prices to their highest possible level, it commoditizes inventory such that impressions that could be packaged up and sold for a justifiable premium go for an absolute song. This is bad in the short, medium and long term for publishers and advertisers.
There are very few markets where an auction model favors the buyer and the seller at the same time. It favors the seller only when scarcity and competitive demand combine. And it favors the buyer whenever either of those things does not exist. Consider the antiques and arts world. Some lots reach astronomical prices not just because of the scarcity of the object, but because there is at least one buyer competing for that very object at the very same time. On these occasions, the selling price is usually far in excess of the justifiable value of the object sold. Great for seller, bad for buyer.
Conversely, when a lot is either quite ordinary or there’s only a single able and willing buyer, the lot price rarely reaches the same level as its justifiable value. It is by buying these very lots that antiques dealers survive.
Now, let’s think what an antique owner might achieve were he instead to market the product via traditional techniques, that is, package the product up, advertise it to seek willing buyers and agree to a price that works for both buyer and seller. Assuming they are matched for negotiation skills, both leave happy.
We see the same occurring in the auction market for online display advertising. Inventory that one or more buyers want at the same time secures CPMs far in excess of their actual value. These impressions are almost exclusively where a user is suitable for retargeting by more than one retailer or more than one agency buying on behalf of a single retailer or even one agency bidding for exactly the same impression for the same retailer through more than one buying point. Great for seller, bad for buyer.
Meanwhile, for all other impressions, prices fall below what might be achieved were they packaged up and sold to advertisers as part of a campaign, whether directly through the publisher team or part of a content or audience-targeted network buy. Great for buyer, bad for seller.
Given we know that for some inventor, RTB will secure prices far in excess of their justifiable value and for others, far below, the only real question for publishers is whether that happens more often than the latter.
The answer to this question depends almost entirely on the nature of the publisher and its inventory. What we see evolving is an ecosystem where in-house sales teams sell all the top, premium inventory; ad networks the inventory that the publisher finds hard to sell but that works well as part of a content or audience package; and RTB platforms the remnant inventory suitable for high-volume, direct-response campaigns and retargeting.
We operate in an environment where publishers face an increasing number of choices in how to monetize their hard-earned inventory. I fully appreciate that there is a place for exchanges and RTB, but do I think all ad inventory should or will be traded this way? Unequivocally, no.
George Odysseos is director of publisher services at ad network Tribal Fusion.
More in Media
Why podcast companies are investing in AI-generated podcast translations despite questionable quality
While podcast networks like iHeartMedia, Spotify and PodcastOne have publicly announced plans to debut AI-generated audio translations, few have gone live yet.
Legal pressure on AI companies illustrates the myriad challenges for companies that want to use or build generative AI tools.