Skip to main content

Guest Blog: The Four Pillars of Data for Programmatic TV Advertising

As the television advertising industry rapidly transforms — changing both viewership and measurement trends and bringing about the rise of new programmatic technology players and data brokers — there’s a lot of hand-wringing about what it all means and what business models will emerge. As we think through the questions, we need to carefully consider the foundation by which programmatic TV will evolve, and how to align our investments accordingly. 

Programmatic TV cannot be defined as simply the automation but the migration from a spot driven marketplace to an audience based market flow. This transition is only possible when complete, actionable data becomes available and the dynamics by which this data is available are equally as important to the transition. In order to succeed at scale, programmatic TV advertising has to be built also on a premise that redefines the outdated manual process of transacting inventory and replaces it with an audience-based model that mainly transacts audience impressions. 

This much-needed shift requires the following:

  • Reliable, timely, and actionable quality data that is certifiably secure and privacy compliant.
  • Once data is available and well understood, stakeholders redesigning their end-to-end transactional processes — related to audience and impressions — based on the underlying data sets in an intelligent and coordinated manner. 
  • Automation of the carefully designed processes will then take place. Simply buying and selling in an automated way does not represent a sound business model in the long run, nor will it keep up with the inevitable changes within the industry.

What constitutes quality data that is also actionable are four key pillars: 

  1. Audience data. The first major shift will be the transition from the outdated metric of age and gender to more accurate definition of audience. With the availability of thousands of attributes, we will use actual and reliable audience data to create segments based on demographics, geographic and propensity to watch a program or to buy a brand. These segments will be created in real--time or pre-defined in advance for each marketing campaign. For example, an automotive marketer who is promoting an SUV would purchase the specific segment relevant to their type of vehicle — those with a determined propensity to purchase SUVs — and would obtain the accurate segment count prior to determining campaign parameters. 
  2. Viewership data. There are no technical or operational hurdles left today for implementing census level measurement in near real-time. Any excuses will be short lived as marketers mandate accurate and accountable measurement of true viewership and of advertising “across mediums” (linear, time shifted, and on demand). Aggregated, de-identified census level viewership data will also unleash the large potential of un-rated networks that reach valuable audiences and deliver prized impressions to programmers and advertisers. When added to the ability of creating audience segments, this type of viewership data will accurately measure reach and frequency “by segment,” impressions “by segment,” and — most importantly — measure conversion rates on the back end “by segment.” This pillar becomes a critical component for marketers to determine their campaign parameters before embarking on analyzing inventory availability and optimal pricing as they evolve to programmatic buying.
  3. Inventory Data. Once a segment is created (e.g., international travelers) and its corresponding viewership is analyzed and determined (e.g., affinity for these 150 programs and 65 networks), the ability to purchase optimized impressions (linear or addressable) will be dependent on the reliability of the information of the underlying inventory; hence the need for near real-time access to inventory data. Anything shorter than near real-time information will be disruptive and cause confusion, costly over/under selling of impressions and definite margin losses to both the buyers and sellers. 
  4. The most up-to-date pricing. Inventory data will require accurate and near real-time pricing as well. The two pillars combined will enable buyers and sellers to take advantage of the laws of inventory supply and demand— especially in a world of faster paced transactions and audience-based buying. Taken together, they will allow for an unprecedented flexibility and value to stakeholders—imagine flexible purchasing and finding the right audience where prices are amenable to the buyer’s budget. 

Each step of the workflow requires unique data and these four pillars of data are critical to the next frontier in television. They will usher in an age where TV advertising becomes less about estimating and more about knowing, less about units and more about audience impressions. Without these four pillars, automating processes will deliver only incremental change. Today’s challenge will be creating the best platform that can accommodate all four dimensions of this data puzzle. What’s needed are solutions that will employ the best and most precise data available so that marketers can optimize ROI on their spend seamlessly. Only the companies that embrace these changes — and especially the ones who are in a position to solve those complexities — will thrive. 

Paul Haddad is senior VP and general manager, advanced data analytics, Cablevision Media Sales.