Analyze This: How Granular Should We Get?

In recent years, commercial ratings have come to the forefront of the industry's focus - so much so that the currency for much of the national television marketplace has changed. But the reality is that while our data has changed, our understanding of commercial delivery has not.

The old Nielsen ratings are program averages of all minutes within a single telecast on a live-only viewing basis. The new Nielsen ratings are program averages of only national commercial minutes on a live plus three-day time-shifted basis. Additionally, companies like TNS and TiVo are attempting to measure television viewing on an even more granular basis.

But the industry migration has been more of a rush to create a data supplier pecking order than to adequately reflect viewing of commercials. This has led to the limited marketplace analysis of commercial viewing behavior amounting to retrofitted reports initially designed to highlight program performance. In short, they are applying program rating analytics to commercial rating data - which just doesn't work.

Is it surprising that "House" had the highest average commercial rating in June? Absolutely not. "House" had the highest commercial rating because it had the highest program rating - the series aired its season finale in June while most other shows were repeats.

When analyzing program viewing performance the first question is, "How many people watched?" If this was a highly relevant question for commercial viewing, the "House" commercial rating standing would have some meaning. But with commercial viewing, the main question becomes, "How many viewers did the program deliver to the advertising and when?"

That question actually has two parts. The first, "deliver to the advertising," acknowledges the fact that people tune in to watch programs, not commercials. The "and when" portion recognizes that programs can deliver very different viewing levels to advertising at different times within a telecast and, most importantly, in proven, predictable ways. In fact, it's that "and when?" that makes Nielsen's commercial ratings only a stepping stone to true commercial viewing measurement and allows considerable room for competitors like TiVo and TNS to operate, as they work off second-by-second level granularity.

Unless the marketplace analytics change to reflect commercial ratings realities, there will be minimal change to our business. And this scenario would be a waste of all the time and money that has gone into, and will continue to go into, the commercial ratings migration. But changes must take place before we'll see true impact from commercial ratings. First, sales and buying systems need to be updated to offer more flexibility. Second, Nielsen also must standardize its data and offer it in the media information type format - the industry standard - for these systems to use. While currently some companies have the ability to aggregate and use Nielsen's minute-by-minute respondent data to create minute-level MIT-like data, many organizations don't have this option. To move forward, we need a standardized industry solution. Third, newer services like TiVo and TNS must become nationally representative to truly impact the marketplace. This paints a potentially rosy picture for a more robust TNS service, while TiVo has no chance as a stand-alone service and will need to identify an operating partner to succeed in the future.

The value of commercial ratings going forward will depend on the level of granularity chosen. It's clear we need a level of granularity below program averages. But how far do we go? Do we need second-by-second data? Is minute-by-minute data adequate? What about pod averages? Pod ratings and minute-level ratings are predictable; second-level data is new and very limited but appears reasonably predictable too. All three types of data are viable options long term. In all scenarios, historical results can be used to predict future performance. Most importantly, all shed light on how many viewers the program delivered to advertising and when.

In the end, the answer to the granularity question will be as varied as the entities that end up using the data. The days of one provider of television ratings with the one-size-fits-all concept are over. We don't need one solution. We need options in these data and the analytical thought to truly transition from program to commercial ratings.

John Spiropoulos is vice president, group research director, at MediaVest. (

Next story loading loading..