• by Charlene Weisler , Op-Ed Contributor, April 18, 2012

Pat Dineen, SVP Nielsen, is a member of Nielsen’s Local Television Audience Measurement Product Leadership team. His responsibilities include leveraging Nielsen’s technology and know-how in the local measurement sector. Pat is relatively new to Nielsen, coming from CMR, where he built large databases and used analytics so advertisers could maximize their local presence.

In my interview with him, Pat talks about Nielsen and STB data measurement, local targeting and segmentation, Nielsen’s Code Reader, diary measurement and data ownership, as well as offering some predictions of what’s next in media measurement.

Below is an excerpt of the interview. See the rest here.

CW: From your perspective, what are the challenges of set-top-box data for measurement?

PD: First of all we should know that we are at the very early stages of set-top-box measurement and its impact on this industry. And with that comes all of the growing pains associated with a growing technology.

The biggest issue we have is that it is scarce. Less than one in five of the subscriber-paid television households provide return path data that is commercially available. That is too little, in our view -- certainly too little for a good census measurement -- and a challenge [for] how we can effectively produce a representative sample.

The data is, to use a technical term, “splotchy.” It arrives based on the geographies of the various contributing MSOs, and the delivery technology is not standardized across all MSOs. So we wind up with some markets where some neighborhoods are fully covered, and others are not.

We then have the issue of how we can best balance that out. How can we create an effective representative measurement when you only have some neighborhoods providing new data, and others not? So our challenge is how we can leverage our assets, leverage our local panels and integrate set-top-box data with that in order to be able to create representative measurement. And that is exactly what we are doing now with our hybrid measurement methodology.

CW: I am curious to know how you are hybridizing the data in the three different types of markets – whether it is a people meter market, a household meter market or a diary market.

PD: The core of our measurement is our people meter panels -- both our national panels as well as our local people meter panels that are available in our largest markets. The process of hybrid measurement is to leverage that data in a statistical process, where set-top-box data can provide you with the large sample size you need in order to take away some of the sample size issues associated with smaller panels.

But your panel helps you overcome the challenges associated with the non-representative nature of set-top-box data and where it comes from. There are a couple of different techniques we are using in different types of markets.

We will be presenting the details of that to our clients, as well as to the MRC very shortly.  But the bottom line is, how do you take the challenges of a small panel away and take away the challenges of a large data set that is skewed by where it comes from.

CW: Is there a priority to concentrate on the diary-only markets first?

PD: Not necessarily. Each one of the market segments poses different challenges and have different client needs. And we are subject to the availability of data -- what markets are delivering set-top-box data. So all of this has to be taken into account when deciding which markets will get our initial attention for hybridization.