Make Your TV Ad Dollars Work Harder

September 24, 2019 Robert Zepf, Director, Marketing Effectiveness, Nielsen

Every day, marketers, media buyers and content producers are faced with an increasing range of platforms, channels and devices. Deciding which ones to spend their advertising dollars on is complex.

TV remains one of the few ways to reach a mass audience, especially for tentpole events such as the Superbowl and the Oscars. But the ecosystem has changed dramatically, as viewers now choose among linear TV and a host of on-demand options. 

Connected TV (television programming delivered via over-the-top devices) has changed how consumers watch, transforming the experience from appointment-based viewing and recording to one in which viewers watch what they want, when they want, wherever they want from any of hundreds of different OTT apps and services. The number of households using only over-the-top streaming services and devices has tripled in the last five years, according to the Video Advertising Bureau.

TV still commands a significant portion of many brands’ advertising budgets. But getting the best results is complex: there are numerous factors that determine a TV advertisement’s success including creative, frequency, duration, and reach. How long an ad should be, what time slots to place it on, which channels, and how many ads to run are just a few of the options. This makes it even more critical to have accurate data and insights to improve effectiveness. 


TV advertising, used to build brand awareness and create demand, claims a significant percentage of most advertising budgets—almost 32% according to eMarketer. Global TV ad spending on linear TV — real time and delayed viewing — inched up 1% in 2018 to $140 billion, according to WARC. 

Until now, advertisers have not known exactly which variables drive ROAS for TV. Many brands rely on Gross Ratings Points (GRPs) and Target Ratings Points (TRPs) based on audience size and demographics alone. ROAS, a marketing metric that measures the efficacy of an advertising campaign, allows brands to evaluate which tactics are working so they can improve future advertising efforts.

Advertisers need to be able to identify which factors contribute to conversions and drive business. Conventional wisdom suggests that “creative is king” and that boosting ad quality is the best way to achieve lift. Until now, there has been little research into how variables influence outcomes. 


WarnerMedia Ad Sales partnered with Nielsen to help diagnose the impact of TV advertising decisions on return. It wanted to identify the influence of specific variables to guide advertiser optimization of future creative, scheduling and placement.

As the largest marketing mix modeling (MMM) provider in the world, Nielsen is uniquely positioned to answer this question. Using our rigorous marketing mix modeling methodology, Nielsen analyzed the results of studies for 130 brands in 11 different product categories to tease out the impact of TV advertising factors on driving ROAS. This representative sample provided complete data for all metrics. The study analyzed $5.2B in TV ad spend. 

The categories included CPG and non-CPG brands such as household cleaners, cosmetics, personal hygiene, pharmaceutical, snacks, food, beverages, financial services, communications and technology, automotive and retail.

Nielsen tested two types of variables: short-term media factors and long-term brand factors. Short-term media factors included creative, ad duration and quality, daypart, reach, frequency and others. These factors are controllable. Brands can influence TV ROAS through tactical and strategic decisions about the duration and content of ads as well as who, how and where to reach consumers.

Long-term factors included brand size, revenue attributed to media and revenue share. These are brand characteristics and market dynamics that cannot be controlled in the short term, but are still strong drivers of TV ROAS.

The analysis revealed several key findings:

  • TV remains an efficient advertising vehicle, especially for a mass audience.
  • While creative is important, it’s just one of several factors driving conversion.
  • Brand factors such as size and market share play a larger role in ROAS than previously understood.
  • TV advertisers in general have learned their lessons well and are achieving healthy ROAS across categories.
  • Brands of all sizes can optimize short-term factors for greater effectiveness, although lifts will grow more slowly for smaller brands.


Advertisers face an increasingly competitive environment in which distracted consumers use multiple devices simultaneously, including while watching TV. In the digital era, TV is one of the few channels that enables brands to reach a mass audience. 

Because CPG and Non-CPG brands continue to invest a significant portion of their advertising budgets in TV, it is critical that these advertisers have accurate data and insights to improve effectiveness.

Marketing mix modeling is a critical tool that allows brands to understand the efficacy of their advertising campaigns. Nielsen and WarnerMedia Ad Sales are committed to helping brands evaluate which tactics are working so they can improve future efforts and achieve long-term growth.


To get the complete findings and recommendations, download a copy of our ebook: Driving TV ROAS: How to Make Your Next Dollar Work Harder

Request a Demo

Learn how Nielsen can help you measure and optimize your marketing and advertising: Request a demo today.

Subscribe to our Blog

Get news and information in your inbox every month: Subscribe

Previous Article
Clear the Way: How Automakers Can Use Measurement to Zoom Ahead
Clear the Way: How Automakers Can Use Measurement to Zoom Ahead

Auto sales in the U.S. are declining. Learn how automakers can use measurement tools to improve marketing a...

Next Article
14 Ways to Win the Always-On Consumer
14 Ways to Win the Always-On Consumer

Here are 14 takeaways for the measurement strategies and tactics marketers need to be successful in 2019 an...



Error - something went wrong!