Everyone working in digital media and ad tech has talked about or heard conversations about “Remnant” and “Premium”. I hear these terms bandied about every day by sales teams, ad traffickers and people working for ad platforms. I think “Remnant” and “Premium” are concepts that have expired their usefulness. It is time to move on. This blog post will explain why… as well as comment on what would be more important!
To start with lets explore what publishers mean when they talk about “Remnant” and “Premium”. If you Google “Remnant” it is defined as is the inventory publishers can’t sell directly to advertisers and is sold more cheaply to market intermediaries. Whereas “Premium” is the inventory sold directly to advertisers and is generally sold at a high price. I broadly agree with these definitions.
However, I have noticed that some of the people talk about “Remnant” and “Premium” differently. Instead, they use “Remnant” and “Premium” as categories of value, i.e. how much it would cost to buy the inventory. With this way of thinking the “Premium” category is the most desirable (above-the-fold placements, special projects, take-overs, etc). By contrast, the “Remnant” category becomes the least desirable, i.e. the Cinderella of the selling stack. Historically publishers threw this Cinderella inventory to the networks since it is hard to sell.
To some, it may be a subtle difference, but any confusion over the value of inventory is great for people playing the arbitrage game. Buy as Remnant and sell as Premium! Google became even richer playing this game.
Programmatic trading has changed things. Publishers now tend to sell “Remnant” in the real-time open auction markets and sell “Premium” in the ad serving stack. Arbitragers needed to find another way to arbitrage and advertisers are exposed to CPM higher rates. Advertisers are exposed to higher rates because they are either buying this “Remnant” via a DSP using third-party data or they are buying “Premium” in the ad serving stack using Publisher first-party data. The problem for people trying to navigate the market is that the terms “Remnant” and “Premium” are not used properly and are not really very useful concepts in the first place. Which one is more valuable in terms of total revenue is also very subjective and depends on the context (e.g. site content, relationships with advertisers, the outlook of the media owner, etc.)
My personal view is that “Remnant” and “Premium” have outlived their usefulness. These terms did not always exist and they do not have to endure. The fact we even have terms like “Remnant” and “Premium” are more a commentary on the fact that there is, in general, an over-supply of inventory, so it is a way of grading inventory in terms of how hard it is to sell. Instead of these dated and misleading concepts to describe different types of inventory people should talk about: target audiences; market data (transparency); disparity between the huge volume of supply and the relative paucity of demand.
Regardless of definitions, the big problem still remains: how should a publisher play his cards when planning inventory monetization? Of course, this depends on the specific situation of the publisher but one big thing that is often overlooked is ad server capabilities! If an ad server is able to dynamically adapt the pacing and priorities with respect to audiences it can be a game changer for a publisher’s ability to control the market for their own inventory. Where a publisher can influence the type and supply of inventory, the concepts of “Remnant” and “Premium” are redundant. Of course ad servers need to be flexible to play this complex game of programmatic trading. Simple and static rotation, capping and pacing are not enough. This also means that humans have to play their part too! Ad trafficking teams need to be able to configure and continually reconfigure the ad server and SSPs and exchanges. They need to up their game and override automated systems where they are not performing.
1. “Remnant” and “Premium” are legacy concepts… other strategies/models can offer better insight and results for a modern publisher:– DMP audiences, dynamic bid floor, guaranteed, private deals vs open market approach etc;
2. Ad servers can help publishers control the market for their inventory with a more sophisticated/dynamic approach to supply which can lead to a serious uplift in revenue. Dynamic inventory allocation and ad serving real-time decisions combined with open RTB interfaces can be available options when people understand the power of inventory control;
3. I didn’t mention it earlier but a single-decision-point-layer in terms of pricing and bid floor control combined with guaranteed delivery is a strategic game changer that publishers can incorporate into their systems. Optimising for CPM with multiple ad exchanges is never going to happen without effective mediation on both an individual impression delivery level and a holistic ad serving level. Effective mediation can be a part of yield optimisation and decisioning for the publisher. The problem is that both “yield optimisation” and “decisioning” are relatively new concepts and most publishers do not fully understand there are no simple answers: there are always trade-offs.