Heavy Lifting

Saturday, April 12, 2008

Interesting radio revenue data from Charlotte

Today's Charlotte Observer has an article discussing the revenues generated by the top 16 radio stadions in the area. Specifically, the revenues are projected to remain flat, which is better than the 3-5% reduction expected to occur in the other markets around the country.

The article provides a table of data, which is all too rare nowadays, which reports the annual advertising sales, the ratings ranking, and the station's format. I augmented the paper's data using the Nov 07 - Jan 08 AQH (average quarterly hour listenter) ratings from Arbitron.

One interesting pattern is the distribution of station formats:


            format |      Freq.     Percent        Cum.
-------------------+-----------------------------------
adult contemporary |          4       25.00       25.00
      classic rock |          2       12.50       37.50
           country |          2       12.50       50.00
            gospel |          1        6.25       56.25
           hip hop |          1        6.25       62.50
         lifestyle |          1        6.25       68.75
              news |          1        6.25       75.00
               pop |          1        6.25       81.25
              rock |          1        6.25       87.50
            sports |          1        6.25       93.75
            top 40 |          1        6.25      100.00
-------------------+-----------------------------------
             Total |         16      100.00

Wow, one fourth of our stations are "adult contemporary?" I have mentioned that the Charlotte area is a wasteland when it comes to interesting radio, and this about shows that definitely. In Dallas there were three talk radio stations (570, 660, 820) whereas here there is only one (1110). Moreover, in Dallas there were three sports talk stations, here there is one (although it broadcasts on two different frequencies). Granted, Charlotte is only about 25% the size of DFW, and that is probably what's driving the lack of selection around here (thank goodness for the internet).

I scattered the annual advertising reveneu against the AQH (which might vary slightly with seasonal changes in demand):

I like the positive relationship between ad revenues and listeners. I find it interesting that there are stations with the same level of listenership with dramatically different levels of advertising revenues. Assuming that the stations have equally competent ad salesman, this suggests that people who listen to certain radio stations a) respond less to advertising, b) respond to advertising but don't spend as much when they do, c) those who advertise don't gain as much from the advertising as on other stations (perhaps they are in more competitive markets?).

This intuition might be supported by looking at the table of advertising revenues, ratings, and format:


     | aqhshare   adsales               format |
     |-----------------------------------------|
  1. |      1.6       3.3               sports |
  2. |      1.6       2.8   adult contemporary |
  3. |      2.3       2.5         classic rock |
  4. |      2.7       4.3               gospel |
  5. |      2.7       9.9            lifestyle |
     |-----------------------------------------|
  6. |      3.2       5.7               top 40 |
  7. |      3.4       5.5                 rock |
  8. |      3.8         9         classic rock |
  9. |      4.4       7.3   adult contemporary |
 10. |      4.6       7.5                  pop |
     |-----------------------------------------|
 11. |      4.8      12.4                 news |
 12. |      4.9       7.5   adult contemporary |
 13. |      5.3       7.2              hip hop |
 14. |      5.7       7.2              country |
 15. |        6       8.4              country |
     |-----------------------------------------|
 16. |      6.4         9   adult contemporary |

The two least listened to stations, one of which is sports talk and one of which is adult contemporary, have the same ratings but $1 million difference in advertising revenues. If the number of commercials offered on the two stations is roughly the same, then the folks who listen to sports talk respond to advertising moreso than those who listen to adult contemporary, or the firms that advertise on sports radio are in less competitive markets where they are willing to pay more for advertising, ceteris paribus.

The most glaring difference is #4 and #5, which have the same listenership ratings (2.7) and about a 100% difference in advertising revenues. Evidently people who listen to Gospel radio aren't down with the advertising, or those who advertise on gospel radio are in more competitive markets where they are not willing to spend a lot for airtime.

The big news-talk radio station has an average AQH ranks it 11 out of 16. I am not surprised because the station plays local talk radio shows all day and UFO radio at night. There are four hours a day when people listen to that station - 12-3pm for for Rush Limbaugh, during which I am sure local businesses are willing to pay top dollar to access, and the midnight hour, during which the station airs the final hour of the Neal Boortz show. The rest of the day, my guess is that not too many people are listening and thus their low AWH.

Just for giggles, I estimated the listener elasticity of advertising sales controlling for "format specific" effects:


. xi:areg lnsales lnaqh ,absorb(format)

Linear regression, absorbing indicators                Number of obs =      16
                                                       F(  1,     4) =   15.93
                                                       Prob > F      =  0.0162
                                                       R-squared     =  0.8963
                                                       Adj R-squared =  0.6111
                                                       Root MSE      =  .28901

------------------------------------------------------------------------------
     lnsales |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
       lnaqh |    1.03901   .2603161     3.99   0.016     .3162564    1.761763
       _cons |   .4894813   .3444966     1.42   0.228    -.4669947    1.445957
-------------+----------------------------------------------------------------
      format |          F(10, 4) =      1.099   0.506          (11 categories)

The simple regression model suggests that the listener elasticity of advertising sales is not different from one. That is, a one percent increase in listenership yields a one percent increase in advertising revenue.

Stata Data File

Labels: advertising, econometrics, economics

posted at 4/12/2008 05:06:00 PM (0) comments

Comments: Post a Comment