The Future of TV Ad Planning: Audience or Index-Based?

An interview with Kyle Hubert, the chief technology officer at Simulmedia

By Matt Collins


Kyle Hubert courtesy of Simulmedia

Some innovations catch on faster than others. Washing machines, color TVs, and smartphones each presented such clear advantages over the devices they replaced that it's hard now to imagine life without them. Other innovations have required more time to establish their dominance. Consider the internet. The first message delivered over its earliest iteration was sent in 1969. It took more time for graphical user interfaces and bandwidth to enable the internet we know today.

One of the most interesting and innovative changes in the world of TV advertising involves the way marketers target audiences. Even though many brands now have a sizable database of customer information, nearly all their TV media is against indices and standard Nielsen demos, mostly adults aged 18 to 49. But there's a new and, in many cases, demonstrably better approach called audience-based buying. It has evolved into a credible way to unlock the value of the customer data companies have acquired.

Deployed properly, audience-based targeting and planning allows brands to bring down both their media costs and their cost per customer acquisition. As with the internet, though, adopting this approach requires just the right combination of inputs, including technology, data, and talent. Fortunately, those resources are more available today than ever, making audience-based TV advertising accessible to just about every brand and agency.

In light of this option, marketers should know what method their plans use today and then ask themselves if there is a chance an audience-based approach could save the company money or generate better results.

What should brands do? Should they change some, all, or any of their TV targeting? Here's what Simulmedia's Chief Technology Officer Kyle Hubert has to say about it.


Q. What does an index say about the makeup of a TV audience?

An index tells you the relative difference of an audience's composition within a viewing event compared to the average audience size in the television universe. If 17 percent of everyone in the TV universe is within the target a marketer would like to reach, and a program has an index of 140, that means about 24 percent (or 140 percent of 17) of the people watching that program are members of the target audience.

Really, an index is a way for marketers and their agencies to evaluate the value of a program against their target audience. It should be noted, though, that a higher index doesn't always correlate to the best purchasing decision, because a lower-indexed program may be priced more favorably on a target-audience reach basis. Because of this, agencies typically consider additional factors like cost, context, and viewership when ranking the value of an inventory unit.

We don't use indices because they can only tell you how high the concentration of your target audience is likely to be within a given program. And, it tells you nothing about the overlap of the audience between two programs. So even if you buy two high-indexing shows, it's likely that you'll be exposing some of the same viewers to your message in both instances, thus boosting your frequency of exposure rather than extending your reach to those who haven't already seen your ad.

This wasn't an issue when there were fewer networks, so everyone bought that way. Most still do because that's the way their systems are designed. In today's age of audience fragmentation though, there are literally hundreds of billions of other media plan options.

With numbers like that, there's no way a person can develop the optimal media plan without technology to help. For a marketer to maximize reach against the target audience within a given purchasing cycle requires a fundamental change in how spots are selected and purchased for a media plan, moving from a relative composition index to a true audience-based method, where the viewing habits of specific audience members are accounted for.


Q. Can you predict if a person is going to be watching a certain show?

Yes, with statistics and machine learning, we can define audiences using tens of thousands of different attributes and create precision media plans for strategic segments like pet owners, yogurt buyers, or frequent travelers.


Q. That's a big part of audience buying; can you explain audience buying versus index buying?

True audience-based buying does not account for the relative composition of an audience, or the context within which an audience is likely to be found. Instead, it values the raw number of individuals in a target audience who actually watch a given program, their likelihood of being exposed to an ad, and the cost of reaching them with a particular spot.

Basically, an audience-based approach enables advertisers to buy people rather than programs.


Matt Collins is the SVP of marketing at Simulmedia. You can email him at




Comparing an Index-based Plan Against an Audience-based Plan

Simulmedia created two media plans targeting a sample audience of women ages 25 to 54 who are cooking enthusiasts and have kids. One media plan was created using an index-based approach, the other using an audience-based method. Each campaign was given a budget of $500,000 for a one week flight.

The index-based plan included inventory selected to maximize the target-audience index of the campaign schedule according to Nielsen data, and was subject to constraints of budget and minimum network spending levels. The audience-based plan included inventory selected to maximize the target audience reach subject to the same constraints.

The comparative analysis was designed to demonstrate the different outcomes when planning is based on composition index versus audience reach. Recognizing that in actual TV campaign planning, composition index is one of several factors considered in assembling a campaign schedule, spots for the index-based buy were selected on index alone— without respect to other factors — to provide a truer comparison between the different methodologies and gain insight into their relative values.

Here are the results:

The inventory selected for the index-based plan had an average index of 188, or 32 percent higher than the average index of 127 on the audience-based plan. Yet the audience-based plan reached nearly 40 percent more of the target audience in total and 48 percent more than the index-based plan.


How the Index-based Plan Performed Against the Audience-based Plan

Index-based Plan
Audience-based Plan
source: Simulmedia internal data

The other data point that immediately jumps out is the CPM. The audience-based plan has a target-audience CPM of $55.75, which is 32 percent lower than the index-based plan CPM of $82.84. What do these metrics reveal? That optimizing for unique reach and efficiency, rather than index, helps marketers do more with each TV advertising dollar.

An index-based plan will tend to fall on a fairly narrow range of networks. In this case, the index-based plan included 239 spots across 23 networks, but just three of those networks — VH1, MTV, and TLC — accounted for 45 percent of the impressions delivered.

The audience-based plan selected 512 spots across 54 networks. That's more than twice the number of spots and networks found in the index-based plan. The media weight was more evenly distributed across the plan, with only two networks accounting for more than 10 percent of the total impressions.
— M.C.