“Data Driven Thinking” is written by members of the media community and containing fresh ideas on the digital revolution in media.
Today’s column is written by Eric Picard, CEO at Rare Crowds.
The industry is rife with conversations about programmatic premium media buying and selling right now, and I could write a column about each of these things, and I could write both an opposing and supporting statement about each. But let me give you one sentence on each from a pro and con point of view:
Programmatic Premium is about increasing efficiency of buying and selling.
Pro: Efficiency gains in buying and selling media are critical – we’re 10-15x less efficient to buy and sell than traditional media.
Con: Increasing efficiency ,while a solid goal, leaves a lot of value (and money) on the table.
Programmatic Premium is not about RTB.
Pro: Real-time bidding isn’t the only way to handle programmatic media buying and selling, and a lot can be done by letting buying and selling systems directly integrate – such as DFA and DFP or any of the numerous Programmatic Premium (or Programmatic Reserve) platforms out there.
Programmatic Premium will destroy sales teams.
Pro: Cost of sales is far too high at most publishers today – for each sales person there’s an “army” of account managers, ad operations, and technical support teams that drive cost of sales for digital far higher than other media, and many of these roles are entry-level with high turnover.
Con: Sales will always be necessary and have a role in the media business, and “order taking” is not a value added process – which is what Programmatic Premium will help simplify.
Programmatic Premium will finally kill off RFPs.
Pro: Sales teams hate the RFP process, which are incredibly time consuming to respond to, and rarely pay off – so the idea that Programmatic Premium will kill them is the carrot held out by many in the industry, and in truth RFP process will be vastly reduced if it is adopted.
Con: RFPs are the description of demand (especially when coupled with a media plan) that publishers should be hungry to review – since nothing else they get from a buyer gives them full insight into what value they could offer.
Why efficiency can’t be the only thing we solve
Of all these discussions, the biggest one I want to call out is the idea that Efficiency Gains are the real value of Programmatic Premium. As an industry we are in far greater trouble than we realize – and efficiency is only a small part of problem we need to fix. There is a greater than 100% price drop when content is consumed over digital distribution rather than traditional media models. And the costs of sales are much higher. What we’re facing is nothing short of the implosion of the media economy as consumption moves to digital channels. We need ways to increase the value and price of digital media significantly, and increasing efficiency lowers costs but doesn’t increase value or price.
Here are two mechanisms we can run after that can increase value and price:
- Targeting more valuable audiences that match better against goals
The basis of media theory starts with an ideal target audience that a buyer wants to reach with a marketing message. These ideal audiences are based on market research done by the marketer and turned into a set of personas that the marketer has defined in extreme detail, and then done extensive research to understand their numbers and media consumption habits. They will know for instance that a key persona makes up 10% of the audience of a specific television show, magazine, or web site. This knowledge drives the decisions to buy ads against specific content experiences, and therefore the volume and price that is needed to make that buy worthwhile.
Media buyers create a simple exchange rate – sometimes held just in their heads – of the price and volume requirements as they pit one publisher against another in their quest for audiences. With the advent of targeting in digital media, the theory has been that this targeting data will be used to create better proxies to audiences than content. But there’s a big problem with that.
Current ad platforms can define inventory by no more than 2-3 targeting attributes. They may well have dozens of targeting attributes available – even on an individual impression. But reserving inventory with more than a few attributes is impossible. And since inventory gets small as you layer on more attributes, nobody spends the time to manage buying them in RTB either. But what is ideal is to find highly targeted and therefore more valuable matching audiences for advertisers. And this is an area where innovation is now happening in our space.
- Concentrating Demand against Supply
Paid Search teaches us a lot about how to drive up prices on inventory. It has primarily to do with fracturing media from large undifferentiated pools of inventory down into micro-marketplaces. In paid search this is done by using keywords to define inventory.
Paid Search micro-marketplaces (inventory) have the following characteristics:
- Low Volume (small number of monthly impressions)
- Uniform (standard ad format, simple definition of the inventory)
- Highly desirable (since paid search inventory tends to exist at the purchase end of the funnel, it tends to be valuable and highly desirable to buyers)
When a keyword (or query) micro-marketplace has optimal demand, it drives the price on that micro-marketplace upwards (creates upward pricing pressure) until an optimal or “true” price is found for that micro-marketplace. In Paid Search overall, the demand tends to concentrate on enough keywords that it drives the overall yield of paid search as a marketplace upwards as it continues to grow. Paid Search has proven that we can create upward pricing pressure on inventory, but it isn’t clear how we can extrapolate from the very small micro-marketplaces of the keyword auction to the relatively huge chunks of inventory in online display.
That’s where Programmatic Premium can provide some help, but it requires additional innovation beyond just efficiency gains.
If we can improve the description of demand – essentially take all the demand data out of the current Media Plan and RFP – and find pieces of inventory that match on lots of inventory attributes – then we can recreate the kind of upward pricing pressure on more inventory in other media types. The idea is to find the intersection of demand with small pockets (micro-marketplaces) of supply.
In order to pull this off, the demand must be fully expressed across many attributes, and the supply must be able to be matched across many attributes. As we gain this ability across the industry, smart matching engines can find places where supply and demand are optimal and can increase the value and the price of inventory at scale. That means prices will start to become comparable to what we see in traditional media, which is how we can save traditional media as we know it.