Skip to main content

Nielsen Media HD household numbers come from direct observation

In late October, Nielsen Media Research published its first estimate for the number of HD households in the United States.

The numbers surprised many because they reflected an HD household number significantly lower than estimates from other sources. Nielsen pegged the number of households capable of receiving HDTV in the United States at 15.5 million and those that actually have the HD equipment needed and receive at least one HD channel or network at 12.7 million households.

The 15.5 million mark is less than half the number of HD households the Consumer Electronics Association (CEA) has estimated. In June, the CEA released figures showing that 32 percent of U.S. households owned an HDTV, or about 35 million households.

With such a wide discrepancy, HD Technology Update thought it might be a good time to ask Nielsen Media how it arrived at its numbers. Anne Elliot, VP of communications at Nielsen Media, provided some insight into the organization’s HD household figures.

HD Technology Update: There is a large discrepancy between Nielsen Media’s recently released numbers for the HD capable households and the number the CEA quotes. Why?

Anne Elliot: This sort of reminds me of a similar question that arises when satellite services give their numbers, and they are based on their subscribers, which don’t match our numbers.

One difference, and I can’t speak to how the CEA comes up with those numbers, but ours are based on our samples, which are based only on TV households. So, if they’re talking about the number of sets sold, or it may be (based on sets) shipped to retailers, ours is based on households.

For example, one household may have more than one HD set. So, if the manufacturers are talking about the number of sets shipped, that’s one difference. Another is because our samples are based on households; we do not include restaurants, bars, hotels and the workplace. I know when I have the opportunity to stay at a really nice hotel now, I’m finding more really nice flat-screen HDTVs. That wouldn’t be included in our number.

HD Technology Update: Could you elaborate on how Nielsen arrived at its numbers for HD households?

Anne Elliot: We held out coming out with a universal estimate until we could do it based on our field reps actually seeing the television. As you can imagine, if you call somebody and ask, “Do you have HD television?” everybody who watches TV and sees “Now in HD” — a certain percent of those people I would guarantee will then assume they have HD. That does not mean they have the equipment to see HD.

Right now, we have only come up with estimates in our metered markets, and at this point our only People Meter markets are national and 13 local markets in which we have People Metering equipment. That is in part to be sure that what we are seeing is a genuine HD-capable television, and then that there are channels received in that home that actually are telecasting HD signals.

HD Technology Update: As I recall, there are about 35,000 households in the United States with People Meters. Could you discuss what it means to be a People Meter household and why that’s important in estimating HD household numbers nationwide?

Anne Elliot: Because we actually go into all the homes to connect all of the equipment, it’s actually observed in the home.

Somebody from Nielsen actually goes in and connects that set to the People Meter. We go into 35,000 homes with the People Meter. At any given time, there is a certain percentage of those homes that we’re going to and saying, “Thank you so much for being a participant; your time has ended.” Then we go out and recruit new homes. It’s a constantly changing sample. There are always people going in and going out, so our HD universal estimates are based on what our field representatives have actually seen and installed in those homes. For the metered sample, we are going into about 35,000 homes.

HD Technology Update: Doesn’t Nielsen also sample outside People Meter markets for TV ratings, and how does that factor into the HD household numbers?

Anne Elliot: At this point, we have only come out with our national sample and our 13 local People Meter samples. We have another 43 local metered samples, but not all of them have been included in this yet. That’s our next step.

In People Metered homes, you can only be in the sample for a maximum of two years. That has enabled us over the past two years to collect all this information for these homes. In set meter samples, you can be in the sample for up to five years. So, in some of those samples, we don’t have enough data as yet to do those samples for HD households. But we will continue to collect data for this, and as that sample turns over, we will be able to collect data for them.

HD Technology Update: What happens if a People Meter home upgrades an SD set to a new HDTV?

Anne Elliot: Part of the challenge we face is maintaining close enough contact with these sample homes without influencing the way they watch TV or purchase a television, but we certainly make it very clear when we first approach them and install them that anytime they buy new equipment they should call us.

For example, if someone from a sample home buys a new TV set and we don’t get out and meter it, we’ll probably pick up a dramatic change in viewing that might trigger us to call them and say, “We’ve noticed there is no viewing in your living room anymore; did you get a new set?” Then we come back out and install it.

We check each day to ensure that the data we get is good and indicates that the equipment is in working order. So, if one of our meters gets unplugged, we have a way to know that has happened. We have a way to contact them if they’ve gotten a new television and we haven’t reinstalled. For the most part, we have terrific relationships. We call them a couple times a year to check up.

HD Technology Update: Is the national People Meter sample made up of the 13 local samples?

Anne Elliot: We do local market (samples) designed to reflect that market. There is a separate sample for the national market to represent the entire United States.

The national sample is not solely a sum of the local markets. It began as a separate national sample. Until a few years ago, it was national. We’ve now brought the sample to 13 markets. It’s a separate sample using area probability sampling techniques.

Tell us what you think!
HDTU invites response from our readers. Please submit your comments to We'll follow up with your comments in an upcoming issue.