Geography Quiz

Which metro area does Harrisonburg Virginia belong to?

+-----------------------------------------+
| MetroArea                               |
+-----------------------------------------+
| Charlottesville, VA                     |
| Fredericksburg, VA                      |
| Harrisonburg, VA                        |
| Norfolk-Virginia Beach-Newport News, VA |
| Richmond, VA                            |
| Roanoke-Lynchburg, VA                   |
| Winchester, VA                          |
| New River Valley VA                     |
+-----------------------------------------+

[I’m pretty sure the correct answer is not Sheboygan Wisconsin which is the current error in the database]

This is the thing that Radio, Nielsen and advertisers obsess on (and the FCC indirectly)

Grouping communities together creates all kinds of mischief. The official methodology used by the FCC is to split the country into metro areas by county, and grouping counties with any significant population together as if they are a homogeneous group. (That is not a transgender reference!)

So right now, I live in Charlotte North Carolina – the city has a population of about 800,000 but the ratings area for the Charlotte Gastonia metro area has 2.2 million people aged 12+. So how far out from Charlotte do you have to go to find ~2.5 million people (including children). Part of why “Charlotte” doesn’t understand why they don’t control North Carolina politics is they think 2.5 million people self-identify as cosmoplitan, diverse, progressive democrats. I can assure you Gastonia is nothing like that, neither is Waxhaw. The “new South” mentality promoted within Charlotte covers no more than a million people, most of them immigrants from elsewhere.

Arbitron (now owned by Nielsen) produced this map in 2012 (very big)
http://www.arbitron.com/downloads/Arb_US_Metro_Map_12.pdf

In the eyes of Arbitron, “Charlotte Gastonia” includes:

Mecklenburg
– Charlotte, Huntersville, Matthews, Cornelius, Mint Hill
Cleveland
– Shelby
Lincoln
Gaston
– Gastonia,
Iredell
– Mooresville, Statesville
Rowan
– Salisbury
Cabarras
– Concord, Kannapolis,
Stanley
Union
Anson
York (SC)
– rock Hill, York
Chester (SC)
– Chester
Lancaster (SC)

Statesville in Iredell County has a population around 25,000. You are counted toward making Charlotte the 24th largest metro area. If you’re running a radio station your audience is Statesville. Your stations are not heard in Charlotte and definitely not in South Carolina. So when advertisers come calling, what do you say? I have a 20% share of the people listening in Statesville or I have a 0.01 share of 2.2 million people in Charlotte-Gastonia?

Starting to see the problems Arbitron caused?

This entry was posted in About the Guide, Radio Biz. Bookmark the permalink.

14 Responses to Geography Quiz

  1. Fred Stiening says:

    You might be thinking the quiz is a trick question…

  2. Fred Stiening says:

    I just bumped into this small market guy’s testimony to Congress
    http://mediaservicesgroup.com/articles/Alex%20Kolobielski%20Testimony.pdf

    https://www.c-span.org/person/?alexkolobielski

    The 4 hour C-Span video will make you puke. John McCain was chairman in 2003 (first W Bush term). Lee Dickey from Cumulus is arguing with McCain about the Dixie Chicks, Rep Menendex (D-NJ) is arguing that Spanish language stations should be treated as a different “market” than English language stations for purposes of ownership limits. A music artist guy argues that radio is essential for music discovery so big radio stifles new artists, etc

  3. Parrott says:

    Hmmm, I would have thought the greater Harrisonburg, Staunton, Waynesboro
    metroplex area would have been one area together? : )

    with reference to the mcclown vid, I could not watch to much. That is chaos with everyone talking past each other. No wonder nothing positive for the working man ever gets done.
    I remember Laura Ingrham mentioning years ago that MS13 was moving into Harrisonburg. I hate that for the local Mennonite community there.
    parrott

    • Fred Stiening says:

      The underlying thought process is that metro areas are defined by the TV stations, not radio. Analog TV and FM radio had similar ranges. In Statesville, WBT-TV came in from Charlotte, but so did the stations from Winston Salem and Greensboro. This was before cable TV and the “must carry” rule.

      Because “DMA” is a highly proprietary concept. To avoid any potential issues, I did my own algorithm, assigning stations to big cities based on their distance, not county boundaries.

      The proposal in 2003 was to allow radio station owners to designate which metro area a station belongs to, enabling manipulation of the ownership limits by moving stations in and out of a metro area – just by shuffling paperwork, not moving the transmitter or changing the programming. Clear Channel and Cumulus have played the game since. In 2003, Cumulus still had not acquired ABC Radio and Citadel

      • TheChairman says:

        In my former life as a marketing data analyst (1990-2000) I had access to proprietary data provided by TV networks, combined with direct-response (inbound call) data from telecom carriers.

        TV & radio networks identify markets (i.e. set ad rates) based on ADI (Area of Dominant Influence) data, and proprietary metrics they assign to broadcast areas and cable systems/subscribers.

        We had 6,000+ ‘vanity’ phone #’s to customize 100’s of TV ads and isolate responses by market, station, time of day, etc. With that call data, detailed info for each spot-ad could be extracted. I could pinpoint almost every response (caller) location using our ad-buy database, call data, and ESRI mapping software. Keep in mind, this was back when most people still used land-lines.

        The order I used was: ADI, ANI, area code, 3-digit exchange, then ZIP. Obviously there was overlap and ‘cross-talk’ within those data fields. If a caller was in “no man’s land” between 2 or 3 ADIs or ANIs, assigning their location to a particular market became rather arbitrary. (This was my own little side project… my client didn’t seem to want such granular geographic data.)

        I still have the database on a drive, as I was able to synthesize proprietary info from ‘meta-data’ the networks provided us.

        • Fred Stiening says:

          I totally understood what you said. When you listen to talk radio, almost all the advertising is response based, not CPM based image advertising. Even the national ads that are not response based can be measured – people don’t hop in their cars and drive 100 miles to visit O’Reilly auto parts in another city. With clustering and one owner controlling 6 radio stations in a market, it doesn’t matter that much what specific station a listener actually listened to. Genres are just flavors of ice cream, not religions.

          • TheChairman says:

            I developed the system from scratch in FoxPro & SQL Server… it also managed print advertising.

            Importing daily call data from Telcos was a pain, because each IEX & LEC sent data in a different file format and/or record layout. Of course, they were all running some variant of Unix, so I had to write individual routines for each Telco to import and normalize their data… eventually we persuaded them to send ascii text, but they still used strange delimiters or failed to sanitize data.

            Another dilemma which sparked ‘infinite loop’ discussions with the client was: “when does the ‘media day’ (or week) end for proper calculation of ROI for each ad-run, spot, or campaign?” In particular, 11 pm – 2 am created a dilemma with national ads, along with Sun-Mon transitions.

            • Fred Stiening says:

              Oh, I hate Monday midnight to 6am with a passion. Back in the good old days, that was the window when engineers could take the transmitters down. Programming directors don’t even know what is on the station.

              Some overnight shows like Red Eye radio are live on early Monday morning, but most are live Tuesday to Saturday AM. Add in time zones and DST induced delays in AZ and HI and things get unworkable really fast. The good news is nobody is listening

  4. Fred Stiening says:

    Everything you wanted to know about ratings

    http://ratings.radio-online.com/cgi-bin/rol.exe/arb_menu_rank

    An interesting exercise is to add up all the stations in a metro area and see how much less than 100% the total is. That difference is the folks who don’t buy ratings – most religious, public broadcasting, educational stations can’t justify the cost. Subscribers see all stations but that costs big money and is strictly confidential.

    Ballpark number is Harrisonburg ratings are showing only 42% total. Tidewater and iHeartRadio are the only subscribers

    Going out 25 miles, the nonsubscibers include

    WEMC FM All Classical Music EASTERN MENNONITE UNIVERSITY
    WXJM FM James Madison University
    WMRA FM (NPR) James Madison University
    WLTK FM K-Love
    WBTX AM Folk / Bluegrass / Ballads WBTX RADIO, LLC
    WHFV FM Catholic English
    WMXH FM No Web Site 105.7 MHz LURAY VA HAYDEN HAMILTON MEDIA STRATEGIES, INC.
    WNLR AM Christian TeachinG CHURCHVILLE VA NEW LIFE MINISTRIES, INC.
    WTON AM ESPN the Boss 1240 kHz STAUNTON VA HIGH IMPACT COMMUNICATIONS INC
    WHTE FM #1 Hits Hot 101.9 RUCKERSVILLE VA MONTICELLO MEDIA LLC

    So where is the missing 58% hiding?

  5. Fred Stiening says:

    If anyone cares about the explanation (not likely), the grouping of stations was done probably 10 years ago using Microsoft access tools. The market rankings change over time, especially the tiny markets. At some point, I transposed the cities in market 276 and 277. I don’t actually use market designation for anything other than internally ranking the importance of individual stations, but Facebook like counts are now a more reliable metric for that anyhow.

  6. Fred Stiening says:

    Pointing to the flaws in ratings, let’s create a grossly simplified model. In my make believe world, there are only two radio stations and two time periods. To get “credit” for a listener, they have to listen to one station for 5 minutes during a 15 minute long quarter hour.

    So our hypothetical market has 100,000 residents. 85,000 of them are 12+. 30,000 of those are employed, and 20,000 are retired. 5,000 of the employed people work at a business in another city

    So in radio period #1, the AQH measures the following
    Station A: 7,500 listeners
    Station B: 2,500 listeners

    When period 1 ends, the audience shrinks as people arrive at work and the audience shifts to retirees, stay at home moms and welfare recipients

    During radio period 2′ the AQH measures
    Station A: 500 listeners
    Station B: 2,000 listeners (swap shop is on)

    So the market wide total AQH is 12,500
    Station A has an average AQH of 4,000 (7,500+500)/2
    Station B has an average AQH of 2,250 (2500+2000)/2

    Station A has a top line share of 8000/12,500 = 64%
    Station B has a top line share of 36%

    You are an advertiser and want your ads to run during Period 2. Station A charges twice as much because they say they have twice as many listeners. Where do you advertise?

    • TheChairman says:

      My first question would be, what is the product or service I’m promoting? As a general rule, based on the parameters you’ve specified, I’d choose station B and run my campaign twice as long and/or increase frequency.

      • Fred Stiening says:

        You’re dancing around the fundamental flaw of radio ratings, which you alluded to. Is your product something that will be bought by commuters driving to a job or a 75 year old female retiree with no disposable income listening to see if she is in the obituaries? All listeners are not equal – and getting any demographic breakdown is pretty futile given the small sample size, especially in a small market.

        Off topic – the woman who kicked my butt in Michigan to take care of myself opened the newspaper one day and found herself in the obituaries. She had leukemia so it would not have shocked anyone – but it turned out to be another person who just happened to have the same name, and it wasn’t a common name. Her daughter started getting “I’m so sorry” phone calls. It was a Tom Sawyer or Herbie Cohen moment, depending on which story you like.

  7. Fred Stiening says:

    Here is another problem that CC used to allude to, but I’ll be more specific – around New York City, LonGuyland is its own market area, but the dominant radio stations are based in New York. If I’m assigned a PPM device, live in Nassau County and ride the LIRR into Manhattan to my job at Goldman Sachs and I listen to WABC at work, does my listening count toward my home market of Long Island or New York City? What if I’m a mobile worker who spends little time at home – airline pilot, long haul truck driver, riverboat captain, movie actor, auditor, etc… what if I’m an out of market streaming listener – I live in North Carolina but listen to WGN in Chicago

Leave a Reply