How do they do it? I’ve always wanted to know.
Here in the UK at least, a representative sample of families is chosen to have some monitoring equipment installed.
Nielsen ratings for TV and Arbitron ratings for radio.
The link below explains how Nielsen ratings work:
For radio, the main system depends almost exclusively on diary entries. Arbitron basically sends weekly logbooks to random households (taking demographics of the region into consideration) in the target audience. There targetees need to keep track of their listening habits and return the diary at the end of the listening period. I think they get a check for $5 if the participate.
There is (or perhaps was, haven’t been in the business for a while) another survey method run by Birch which relies on telephoning homes and asking questions; obviously the stations who do well in the phone survey prefer to tout numbers from the Birch measurement, but Birch is taken less seriously because (1) it is neweer, and (2) it tends to skew younger (younger people in general are more likely to participate in phone surveys.
For the more-respected Arbitron, use of the diaries has led to some change on the part of radio format. Ever wonder why most radio contests are done on Thursday? That’s the day of the week the diaries usually arrive in homes. At one station I worked at, the GM requested we tell listeners the time between songs only once every 20 minutes, preferably during the times between 20 and 25 past the hourm, 35 and 40 past, and 50 and 55 past. The reason for this is if a diary entry says the listener has the radio on for at least five minutes in a quarter hour, the station got credit for the entire quarter hour. So if you write “listened to WXXX, 9:00 to 9:20”, the station got credit for 30 minutes of listener time. Yes, it is that crazy.
I met a fellow once who was a radio network programmer. He told our staff about a visit he made to Arbitron offices in Maryland (?) and pretty much summarized the process as a complete and utter crapshoot; it’s amazing they get any credible numbers out of it. Still, that’s what local advertisers use to set rates, so you have to treat them like gold and never, ever get caught gaming the system (there were specific rules about mentioning ratings on-air; one local station raised a ruckus when they started using a slogan like “66.6 FM WXXX; write it down”. Some saw “write it down” as a forbidden call to specific diary holders and complained to Arbitron, but I don’t think anything came of it).
TV is similar with respect to diaries, but I understand Nielsen also uses devices attached to the TV itself in sample households to get more accurate numbers (I think it’s 1 metered household per 100,000, but don’t quote me on that). During sweeps (November, February, May), the same metering is done on the local level, rather than just the national. Networks basically stunt during sweeps to ensure better local ratings for their affiliates, but will of course take any opportunity to crow about national “wins” during sweeps months, even though they’re similarly metered nationally for 3/4 of the year anyway.
Broadcast outlets pay Nielsen for the service. There are different tiers of ratings. Smaller markets may just get reports during “sweeps,” May, November, with less important sweeps of February and July, while the networks in the larger markets (New York, LA) may pay for overnights, meaning they find out the next day how their primetime programming did from the night before.
It’s this quick turnaround, especially for the networks, which can cause new shows to be yanked off the air in a matter of weeks after airing only two or three times.
Anyone know how (or if) Neilsen deals with Tivo, OnDemand and plain ol’ VCRs? We hardly ever watch anything live these days - even if we want to watch *Lost *when it’s on, we’ll tune in 15 minutes after it starts and watch it through the saved OnDemand feature, so we can skip the commercials. Most shows we don’t even watch the same night they’re on.
Would we be entirely useless to Neilsen because of this? Do networks and advertisers even care about people who watch a show at a different time? I suppose most ad agencies would think we’re not worth it, skipping all those commercials, but we actually do watch some, out of choice - we love the Mac/PC ads, and I’ve absorbed enough of that Audrey Hepburn Gap abomination to know how horrid it is, even on fast forward. (Hey, I wonder if anyone’s thought of making an ad really slow, so that when fast-forwarded, you get the message anyway.)
This information is absolutely fascinating to advertisers. For example, if most viewers watch, say, Lost through On Demand services, an advertiser may choose to sponsor the on-demand version instead of, or in addition to, the live version.
An increase in TiVo or DVR use for some shows might be followed by an increase in product-placement advertising; advertising pays for the show, after all, and if you’re not watching the commercials, you’re going to see advertising during the shows themselves. Or the commercials might be made differently so you’re forced to get the message anyway.
I did ask someone who knows about TV, and he said that there isn’t a way to measure DVR usage beyond directly surveying users and looking at TiVo and DVR sales rates. It is possible for cable and satellite providers to measure what viewers watch, but there have been privacy concerns about that.
Robin
Nielsen is starting to offer another version of reporting, the real time vs. the delay, but until the advertising agencies request it from the broadcast outlets, the broadcast outlets aren’t going to ask for it from Nielsen.
I just find it hard to believe that the manual diary method produces an accurate record anymore.
In TV, given the use of remote controls, viewers commonly switch channels very often (men especially). And in car radios, it’s common for people to punch a button to switch stations every time a commercial comes on. Sometimes several buttons to find one that isn’t running a commercial.
I just can’t see people actually writinmg down all these channel changes in a diary.
Yep, Nielson. Basically it is a polling operation. Some of it is automated, but they also ask you to do diaries, so they know how many family memebers were watching what show, not just what channel the tube was tuned to.
They are very pushy. I flat out told them I wouldn’t participate, as I felt it was my duty to lower the reliability of thier data by self-selecting out of thier sample. Still called me a half dozen more times working thier way up the chain of command. AND sent me the diary and reminder letters anyway.
No, they’re not very reliable. You’ll get people who don’t want to admit that they watch WWF Wrestling, so they’ll write down that they watched Masterpiece Theater on PBS instead.