Skip to main content

The Requirements of a Convergent TV Future

A family watches content on multiple screens in their living room.
(Image credit: Eric Audras/Getty Images)

I’ve had the pleasure of being around the media world since 1979. I’ve had a catbird seat to watch the growth and evolution of cable, the birth of the internet, the advent of social media and most recently the launch of streaming.

I’ve watched the industry, especially over the last 10 years, struggle to adopt new innovations that would have benefitted both publishers and advertisers, such as data-driven linear. Since my days at Turner Broadcasting, the industry has had the analytics to support DDL, especially the means to build highly accurate forecasts of delivery against audiences at the day/date/half-hour level. But these analytic capabilities were hamstrung by data and measurement limitations, specifically the size of Nielsen’s viewership panel, which is a mere 40 thousand households. 

This sample limitations dramatically inhibited the value of data-driven linear. Linear continued to hold the standard for reach in the advertising world, but DDL could only be executed for audience targets more than 10% of the United States population. Even for large targets, representing viewing on long-tail networks was very challenged. While data-driven linear was a step in the right direction, the measurement constraints of the time limited its potential value return.

Fast forward to 2022. 

We are no longer data and measurement challenged. Companies like VideoAmp, iSpot, Comscore and others offer sample sizes in the tens of millions of homes, measured via smart TV ACR or ACR and cable set-top boxes. These companies also provide the integration of linear TV and digital video, being able to integrate digital video consumption or ad schedules. Today’s currency wars demonstrate an industry-wide recognition that multiple and alternative cross-platform currencies are the path towards a holistic view of the consumer, to be leveraged for better cross-platform advertising executions and better cross-platform content insights.

But though we are climbing this data and measurement hurdle, the challenges we face have grown. COVID has dramatically escalated both the rise of streaming and linear viewership declines. The ever-increasing array of content, services, and devices creates a maze of fragmentation – exacerbated by a new demand for multiple currency reads. The impending arrival of national addressable inventory will dramatically change how television is bought, sold, and more importantly, valued. And consumer privacy rights remain a hot button issue, determining targeting capabilities and strategies.

This somewhat intimidating horizon provides the foundation of demand for decision science leveraged for advanced analytics in forecasting and optimization that can cross platforms, channels, and currencies and provide real-time strategy in the face of an increasingly dynamic world.

Let’s stop for a second and think about these challenges, and reframe them as opportunities that are available to us today and in the near future.

Better Measurement

We have a chance to reimagine linear TV selling and buying. Aside from the long-overdue move from broad age/sex measurement to audiences, the reality is that the industry has adapted the process for selling and buying to Nielsen measurement, mainly because of the growing instability in ratings. And while those steps provided some degree of stability, we’ve sacrificed precision and potentially yield. The sample sizes that most of these new currency providers offer should give the industry a chance to reset how inventory is bought and sold, which could benefit both seller and buyer. These sample sizes directly address Marc Pritchard’s recent comments calling for an end to the upfront process due to exaggerated audience estimates and guarantees. With these larger sample sizes, networks will enter the upfront with clearer view of their supply, and not need to worry about massive ratings fluctuations (one reason for inflated selling estimates).

Addressable Yield Management

We’ve got to solve for straddling a world where linear TV co-exists with addressable in linear, through set-top boxes or connected TVs. A typical cable network yield manages roughly 140 thousand 30 second spots per year. Translating those spots to impressions means that if a network delivers 50 thousand households per spot, they will have to effectively yield manage 7 billion impressions. The world won’t be 100% addressable, so part of yield management will be the trade-off between addressable and underaddressable C3 (whatever ad clears where there’s no addressable impression served).

Convergent TV

We’ve got to solve for cross-platform reach and frequency management, across linear TV, AVOD, and CTV. Not only does this alleviate a major viewing frustration for consumers, but it returns an immense value for buyers and sellers alike. Advertisers can mitigate waste by explicitly purchasing AVOD/CTV impressions on viewers with low/no frequency and publishers can package inventory across linear and AVOD channels.

Outcomes-Based Selling

Despite some efforts in the DDL space, the vast majority of linear TV remains transacted on demographics with publishers highly focused on yield and agencies on price. In this way, advertisers have gained an understanding of their consumer behavior value through marketing mix and multi-touch attribution reports, or through insights gleaned from digital channels lower in the funnel. To justify its premium pricing, linear TV needs to align its inventory with consumer behavior -- but do it in a way that reflects what a publisher can do -- optimize the purchase and placement of its inventory for better outcomes.

Moving Forward

Each of these opportunities requires advancement in how the market thinks about forecasting. As an industry, we used to rely on long-term historical analyses to build our forecasts. But today, I’m not sure how much weight we should apply to historical data. The reality is that what people are doing today and tomorrow is far more relevant to how we plan media. Most of the issues I’ve touched on require a huge degree of precision to forecast -- forecasting at the unique ID/person/household level to support cross-platform and addressable use cases.

10 years ago, we watched as measurement capabilities underwhelmed our analytical ability, and the data-driven linear initiative suffered. Today, we find ourselves on the opposite side of the spectrum, with an abundance of data and measurement sources lacking the analytical firepower to allow them to reach full value. Let these industry challenges be a call to action for analytic investment so that we can approach them as opportunities. ■

Howard Shimmel is head of strategy for datafuelX, president of Janus Strategy and Insights, and former chief research officer at Turner Broadcasting.