Improving Local Measurement With Return-Path Data
Of all the potential uses of return-path data, (RPD) local measurement arguably offers one of the top initial values. Recently I interviewed Pat Dineen, SVP Nielsen about his work using RPD in conjunction with Nielsen’s current local panel methodologies. Pat spoke of hybridization of RPD with the Nielsen panel in all three types of local market methodologies – the People Meter Markets, the Household meter markets and the Diary measured markets. This method offers the ability to include Nielsen demographics while expanding the household sample size.
His interview sparked discussion about how other measurement companies use RPD to measure local markets. In this RPD land rush, it is interesting to note the different approaches. These approaches are often based on availability of tuning data and the types of data streams that are fused or matched to it.
Bruce Goerlich, CRO Rentrak, says that his company has been focused on RPD data for local measurement for well over four years. Rentrak currently has 8 million homes in its measurement footprint (1 in 14 U.S. homes) and this data comes from several sources – Satellite (Dish), Telco (AT&T), MidContinent and FourthWall – which Goerlich says is all available via the interface. Collectively this represents 98% of all residential Zip codes and 90% of all commercially available set top boxes.
Bill Harvey, CRO TRA, notes that TRA now has per-market sample size of over 50,000 homes in 10 markets, over 25,000 homes in 18 markets, 10,000+ in 26 markets, 3000+ in 45 markets and over 1000 homes in 70 markets. One case study enabled an advertiser who does all spending in spot TV to optimize the buy through TRA’s Media TRAnalytics® Optimizer. The client ran hundreds of reports with this granular data to decide what media to place in 37 local markets.
Jeff Boehme, CRO Kantar, approaches RPD from both a national and local perspective through the use of 103,000 DirectTV homes for national measurement and thirteen Charter markets for local measurement. But Kantar is not offering a standalone local product. They have made a strategic decision to partner with Nielsen for local measurement. “If you only have RPD you are missing a piece of over the air which is impractical to model. We are working with Nielsen to help them enhance their Local Market Service product line and align RPD with traditional measurement and will have a 1 million panel for Charter available by May of this year." Boehme says, “We believe that audience measurement will be driven by RPD. It doesn’t mean it will replace it but it will augment and enhance it. RPD is necessary for all advanced measurement capabilities for advanced advertising such as addressable, interactive and T commerce applications.”
Because of the increasing number of data sources, return path data is becoming more ubiquitous and is available across all geographical areas in some form. Rentrak’s Goerlich says while Nielsen builds up the data on the county level, Rentrak builds up the data on the zipcode level. According to Goerlich, “Counties are not as granular a level of neighborhoods as are Zip codes. There are approximately 3,143 counties, parishes or independent cities in the United States and there are approximately 32,847 zip codes with at least 1 home in them. There is the potential for more gaps if you design weights around counties rather than Zip codes.”
Even within all the different methodologies and the different sources and levels of data available on the local level, the hope for some national clients is that local return path data from its various sources – such as MSO, Telco and Satco - will be prepared in such a way that it can be holistically collected and merged to form a basis for a national measurement. Stay tuned.