The First Move: A Content Analysis of National Radio Airplay and Sales Charts
National airplay and sales charts were tracked during a 16-week content analysis to offer insight into whether radio airplay would proceed music sales. During the analysis, 52 different songs appeared on the sales chart. Fifty of those 52 appeared on the airplay chart before appearing on the sales chart. Sales and airplay indexes suggested that airplay systematically leads sales. Results were inconsistent however and personal taste in music was seen as a large factor in the sale of music.
There is a belief in the music industry and the business of radio that radio airplay is greatly responsible for the sale of the music it plays (Breen, 1991; Burns, 1996; Kinosian, 1995; Kojan, 1993; Maxwell, 1995; Smith, 1989). A former radio program director wrote from his experiences that "without a doubt, radio was the major influence upon all record sales" (Claude, 1977, p. 105).
This claim has received little academic research, yet it continues to thrive in the radio and music industry (Hesbacher, 1974).
The purpose of this paper is to offer evidence into what responsibility radio airplay can claim, if any, in the sale of the music it plays. The music industry
certainly belives there is a strong relationship as they have invested great amounts of staff and money into radio airplay since the 1940s.
Since the beginning of music radio, the music industry has employed Song Pluggers or Record Representatives in order to acquire more airplay for their songs (Lazarsfeld, 1941). The concept has always been that more airplay will lead to more sales (Baldwin & Mizerski, 1985; MacDougald, 1941; Shallett, 1990).
Radio airplay's claims of responsibility for music sales received the bulk of its academic research in the 1940s, when music radio was a new medium.
Examples include 1941, a year when only Broadcast Music Incorporated (BMI) licensed music was played on the radio due to American Society of Composers, Authors and Publishers (ASCAP) voluntarily kept their music off of radio to protest BMI's receiving of certain royalties. That year, Variety magazine's "Best Seller's List of 1941" was wholly BMI music. Peatman
(1942-43) suggested radio airplay was the reason for the total BMI showing.
In addition, Weibe (1940) provided evidence of a song's number of plays on the radio having an effect on that song's position on popularity charts. Erdelyi (1940) compared airplay and sales chart and found that airplay systematically lead sales.
The general practices of the music industry have changed little since the 1940s. They still uses record representatives in order to gather airplay for their songs believing that airplay will lead to sales (Shallett, 1990). In turn, music radio still has the belief they can make songs hits (Maxwell, 1995). But radio programming and audiences have changed since the 1940s. It is warrented that the effects suggested then be tested again.
In the 1940s, music radio was not locally programmed as it is today (Pavia, 1983; Alexander, 1993; Kelly, 1994). Local radio stations then were little more than outlets for what the major networks broadcast (Wrzesinski, 1979). The programming was centralized with little feedback from the audience and because of that, radio was seen as the popularity opinion leader for new music (Burns, 1996; Lazarsfeld & Field, 1946; Peatman 1942-43). Even studies involving countdown shows, which themselves performed research to collect what people felt were the most popular songs of the day, showed that hit songs had always first received extensive airplay before showing up on the countdown charts (Peatman, 1942-43).
Today, the majority of radio is locally produced and the decisions of what new music to play are made in house (Denver, 1993; Love, 1994; Alexander, 1993).
Programming decisions are made in regards to a certain audience rather than a nation-wide mass.
Even so, music programming today is most often done through one person who, because of experience, is placed in charge of picking the new music that plays on the air (Denver, 1993, Love, 1994). Audience research, which was quite prevalent in the 1960s and 1970s has been relegated to being used after a song has
been chosen for airplay if used at all (Kinosian, 1992; Kojan, 1992). Still, decisions are still being made in a local, rather than national, environment.
Today's local programming, without a great deal of audience influence, still allows radio stations to introduce new music to the audience they serve, as it did in the 1940s (Burns, 1996). Rothenbuhler (1985) referred to this ability as an agenda-setting process.
If this is an agenda-setting style effect, music radio leading public opinion of popular music, does that effect of perceived song popularity then lead to the sale of new music? The research of the 1940s argued that it did.
But more has changed than the localization of radio programming since the 1940s. The sheer number of outlets for music, and thus music influence, has increased greatly.
In the 1940s, programming music on a radio station was not the prevalent format it is today (Lazarsfeld & Kendall, 1948). Radio stations that programmed music continually, like WNYC in New York City, were seen as the exception rather than the rule and thus were able to keep a more mass audience. (Lazarsfeld & Stanton, 1941). If one wanted to listen to music, there were few outlets to choose from.
Today, radio is also not the only medium broadcasting music. Cable television and the advent of music videos have created a sort of radio station being played on television (Baldwin & Mizerski, 1985). These stations, such as Music Television (MTV), Video Hits 1 (VH1), and Country Music Television (CMT), are recognized by music radio as part of their programming community, so much so that the three television networks noted above have their playlists printed each week in Radio and Records (1996), a radio industry trade magazine.
In one of the few recent tests of radio's ability to generate sales through airplay, Baldwin & Mizerski (1985) showed music videos and played music for respondents and then asked the respondents if they were likely to purchase the music they saw or heard. The results showed that video had a greater effect than radio on people answering they would buy the music. However, as Baldwin and Mizerski point out, these results are self reported and respondents were then under no pressure to actually go and buy the music. Still, this suggests that radio's ability to influence sales may have been diminished after 50 years.
The methods used by researchers in the 1940s were sound. It is just that the variables of music radio and sales have changed greatly since then (Rutkowski, 1989). Today the outlets for music, and music sales, are far more plentiful. Also, in the 1940s, music sales were measured in terms of sheet music
(Peatman, 1942-43). People had to, themselves, play the music they purchased limiting who would buy the sheet music. Today the technology and availability of recorded music allows listeners the purchase the music they hear and play it when they want.
Although proprietary research into a relationship between sales and airplay has surely been done, academic research has not followed. This paper will attempt, by using a sound methodology from 1940 measuring actual sales rather than self-reported intent to buy, to provide evidence for or against a relationship between radio airplay and music sales.
The preceding literature review has provided support for the following hypotheses:
H1: A song that sells enough copies to be listed on the national sales chart will have first received enough airplay to have been listed on the national radio airplay chart.
H2: Radio programming will tend to move a song into a higher position on the national radio airplay chart before that song is moved to a higher position on the national sales chart.
Data for this study were collected by following the national radio airplay chart, The Country Top 50 printed weekly in Radio and Records, and the national
sales chart, Country Singles Sales printed weekly in Billboard magazine. The two charts were followed using a spreadsheet charting each song's movements up or down either chart. A Table was created for each song that appeared on either chart during a 16 week period beginning with the charts printed for the week of June 24, 1995. This was the first week the Country Singles Sales chart was published. A 16 week period was chosen because it was enough time to allow a song to enter the chart, mature, and then fall off. Additional chart data from before the 16 week content analysis was collected from the Country Top 50 airplay chart when a sales versus airplay time line needed to be established. Any additional data collected are noted in the individual song Tables. Table 1 is an example of one the of song Tables.
A simple index, "I", was created to measure a time lag between airplay and sales for songs that had a considerable portion of their life cycles during the
content analysis. Songs chosen for this index were on both charts for at least four weeks. I was computed using the following formula:
I = (Wxr)/r
"W" is the number of each week from the beginning of the observation period. This runs from one to 16. "r" is the rank of the song in the corresponding period.
The result of the formula is a weighted average of each song's movement on the chart. A difference was calculated between a song's sales and airplay by subtracting the airplay index from the sales index. The difference was positive when airplay lead sales and negative when sales lead airplay.
The above methodology is the same used by Erdelyi (1940).
A total of 134 songs were tracked during the 16 week content analysis. Of those 134 songs, 52 sold enough copies to make it to the Billboard Country Singles Sales Chart, a 39% crossover rate.
Only 39% were able to be tracked due to differences in the sales and airplay charts. The airplay chart, is driven by music and radio industry forces and tends to move songs up and off of the chart relatively quickly. In contrast, the sales chart is driven only by consumer choices and does not need to move songs off the chart to make room for other newer songs. It is for sure that songs other
than the ones noted on the airplay charts here sold copies. They just did not sell enough copies to place in the top 25 selling songs for that week.
Fifty of the 52 songs that appeared on the national sales chart received national radio airplay before appearing. The two songs that ranked on the national sales chart without showing airplay never ranked on the national airplay chart in this study.
It was possible to find the highest airplay chart position of 39 of the 52 songs that showed up on the Country Singles Sales chart. These are songs whose sales peaked and then began to fall before the end of the 16 week content analysis. The songs rose to a mean peak ranking of 7.3 (median = 4). The breakdown of the peak positions are detailed in Table 2.
It was also possible, for 26 songs, to chart the number of weeks that the song was on the airplay chart before it sold enough copies to rank on the sales
chart. Results showed songs were on the airplay charts an average of 6.4 weeks
(median = 6.25 weeks) before appearing on the sales charts. The lengths of time
a songs was on the playlist chart before appearing on the sales chart are detailed
in Table 3.
Twenty-three of the songs charted met the requirements noted in the methodology section and a sales and airplay index was calculated for each. The song titles, indexes, and difference scores appear in Table 4. In all but one case,
the difference score was positive suggesting that airplay lead sales. The one difference that was not positive resulted in a score of zero.
H1: A song that sells enough copies to be listed on the national sales chart will have first received enough airplay to have been listed on the national radio airplay chart.
There is some evidence to support this hypothesis, yet the results are inconsistent. Of the 52 songs that entered the national sales charts during this content analysis, 50 had national radio airplay before selling enough copies to show up on the sales charts.
The question is then how much airplay is required in order for a song to appear on the sales charts. Table 2 shows the number of weeks a song was on the airplay chart before showing up on the sales chart. The average was 6.4 weeks covering a wide range of time being between three and 12 weeks. This large span
of time suggests that there is something more to why a song receives sales than airplay alone for if airplay was the main factor then the amount of airplay a song
received before showing up on the airplay charts would offer a more uniform range of time.
The two songs that showed up on the sales charts without showing up on the airplay charts were Redneck Stomp by Jeff Foxworthy, and Hog Wild by Hank Williams Jr.
A plausable reason why Redneck Stomp did not show up on the airplay chart is due to its being a novelty song. Novelty songs often will receive airplay but will not be listed on a regular playlists, thus not reported to Radio and Records and not listed on the national playlist charts. The song is played more as a comedy segment rather than playlist song. Sales charts do not exclude novelty songs in the same way and will rank them if enough copies are sold.
The Hank William's Jr. song is more difficult to explain. Hog Wild, is by a
very well known and popular artist, and it only showed up on the bottom of the airplay charts for one week. Even though some radio airplay might have contributed to the sales figures, Hog Wild probably received airplay on television
video shows. Still, there is no strong plausable reason from this study for the song showing up on the sales charts.
Results of this content analysis suggest inconsistencies in the airplay rank a song must achieve in order to have measurable sales. Table 3 shows the playlist
peak position of the songs that showed up on the sales charts. It is true that the majority of the songs, 85%, that made the sales chart also made it into one of the top 10 positions on the airplay chart, but there are exceptions. Songs such as Alison Krauss's Baby Now That I've Found You had substantial sales while only barely making it onto the national airplay chart, peaking at number 42.
This suggests that airplay alone will not sell a record. There has to be an interest on the part of the listener. Still, the results suggest that a song usually needs to receive some airplay before selling enough copies to make the national sales chart.
H2: Radio programming will tend to move a song into a higher position on the
national radio airplay chart before that song is moved to a higher position on the
national sales chart.
There was also some inconsistent support for this hypothesis. Table 4 shows that in all but one case, the index for airplay produced a higher ranking than the sales index. Differences, except one, were always positive suggesting airplay
proceeded sales movement up the chart. A t-test of the two sets of index data showed that the means differ significantly (t=+67.5, two tailed, p<.001). As was stated in 1940, "airplay systematically preceded sales" (Erdelyi, 1940, p. 698).
Examples of the sales charts following the movement of the airplay charts include Tables 5, 6, and 7. Each Table shows a general upward movements on the airplay charts before an upward movement on the sales chart.
These results, however, must be tempered. To state that because radio airplay systematically preceded sales, does not denote a causal relationship. It only offers the suggestion that airplay might have a strong hand in the process.
Is radio airplay responsible for the sale of music as is believed by the music industry and many radio programmers? The answer to the question is that this study does not provide enough information to tell. It might appear that airplay made for sales. And in part, it probably did. But a radio station cannot play any song it chooses and create sales just through a great deal of airplay. Some songs that received a tremendous amount of airplay did not rank on the sales chart. A good example of this is John Berry's I Think About It All The Time. The song reached number one on the national airplay charts yet it did not appear on the sales chart. Other examples include Lee Roy Parnell's, A Little Bit Of You which peaked at number one on the national playlist charts and Pam Tillis', In Between
Dances which peaked at number two. Neither showed up on the national sales charts.
In turn, the opposite is not true either. A limited amount of airplay does not mean low sales figures. Song number 79, Alison Krauss's Baby Now That I've Found You peaked at number 42. This song received sales with relatively little airplay. Hank Williams Jr.'s Hog Wild also received some sales with no recorded airplay.
The above examples are not to say that airplay does not play a role in the process, quite the contrary, but there appear to be exceptions. The majority of songs, 50 of 52, that showed up on the sales chart received airplay before selling. This suggests that a song will usually have to receive airplay, or be presented to an audience, in order for the audience to be made aware of the song, and then sell. In terms of Hank Williams Jr.'s Hog Wild, it is important to remember that even though the song did not receive enough airplay on the radio to make the chart, that does not mean the song received no airplay. As outlined above, there are many other outlets for music than just radio.
It is not difficult to understand why record representatives want songs played as often as possible. The more plays, the more people that are exposed to the music. In addition, continued airplay could reinforce that the song is still popular
and can be purchased. Whether or not the music is purchased obviously deals with more than airplay alone. Gans (1979) wrote, regarding media effects, that
the media simply deliver the message. It is up to the audience whether to act upon that message or not.
Also, time frame in regards to amount of airplay is not the only factor involved. Media effects research indicates that there are also personal preferences and reactions to the music involved in the process.
Iyengar and Kinder (1987) wrote that media effects are strongest on those audience members who have an emotional reaction to the items being presented. People will have emotional reactions to the music they listen to, and those emotional reactions may result in music sales. (Chaffee, 1985; Frith, 1988; Hirsch, 1971).
If the radio is playing a song that creates an emotional reaction within a listener, he or she could make the decision to purchase that song no matter how many times it has been played. In the study by Baldwin and Mizerski (1985), respondents said they might purchase a song after only one hearing. This is also a viable explanation for any of the sales gathered in this content analysis including Hog Wild and Redneck Stomp.
The results of this study suggest that a radio station's airplay can claim some responsibility for the sales of music. Yet the relationship is not consistent. A few exceptions to the generally followed rankings were found. Although the majority of songs that appeared on the sales charts also had high airplay chart rankings, a few songs sold enough to appear on the sales charts with little or no measurable airplay.
Results also suggest that the decision to purchase a piece of music might be influenced as much by the consumer's tastes as by the level of airplay a song is receiving on the radio.
It appears that radio airplay can claim some responsibility for sales. But radio's role in the process appears to be little more than the introduction, and reinforcement, of music to the audience. Once the listener is made aware of a song, he or she will then decide whether or not to purchase that song. Even an overt amount of airplay does not insure enough public interest to generate sales. Some songs that made it to number one on the airplay charts never showed up on the sales charts.
However, results suggest that once a song receives airplay and begins to sell, radio airplay does help in the reinforcement of the song. Indexes showed that
airplay systematically lead sales up the chart. Radio airplay appears to have a stronger hand in the sale of music after an audience decides to purchase the music it heard. These are the same general results found in the studies of music radio in the 1940s. It appears that the variables have changed, but the effect of radio airplay have remained.
Alexander, s. (1993, December 3) Examining station playlists. Radio and Records, 20, 34.
Baldwin, L., & Mizerski, R. (1985). An experimental investigation concerning the comparative influence of MTV and radio on consumer market responses to new music. In E. C. Hirschman, & M. B. Holbrook (Eds.), Advances in Consumer Research, 12, 476-481.
Breen, J. H. (1991, March 23). Radio, music biz aren't married. Billboard, 103, 11.
Burns, J. E. (1996, April). The Creation and Following of Public Opinion: A History of Music Choice in Radio Programming. Paper presented at the Central States Communication Assoiciation Annual Meeting, St. Paul, MN.
Chaffee, S. H. (1985). Popular music and communication research: An editorial epilogue. Communication Research, 12, 413-424.
Claude, H. (1977). This business of radio programming. New York: Billboard.
Denver, J. (1993, March 19). Figuring out your optimum music rotation. Radio and Records, 20, 34.
Erdelyi, M. (1940). The relationship between "radio plugs" and sheet sales of popular music. Journal of Applied Psychology, 24, 696-702.
Frith, S. (1988). Music for pleasure. New York: Routledge.
Gans, H. J. (1979). Deciding what's news. New York: Pantheon.
Hersbacher, P. (1974). Sound exposure in radio: The misleading nature of the station playlist. Journal of Popular Music and Society, 3, 189-201.
Hirsch, P. (1971). Sociological approaches to the pop music phenomenon. American Behavioral Scientist, 14, 371-388.
Iyengar, S., & Kinder, D. R. (1987). News that matters. Chicago: The University of Chicago Press.
Kelly, T. (1994, April 22). New music takes over. Radio and Records, 21, 32.
Kinosian, M. (1992, April 17). High stakes dictate conservative tactics. Radio and Records, 19, 47.
Kinosian, M. (1995, January 5) Time for us to flex our muscles. Radio and Records, 22, 50.
Kojan, H. (1992, April 3). Researching new music. Radio and Records, 19, 42.
Kojan, H. (1993, January 29). How PDs choose rotations. Radio and Records, 20, 35.
Lazarsfeld, P. F. (1941). Audience building in educational broadcasts. Journal of Educational Sociology, 14, 533-542.
Lazarsfeld, P. F., & Field, H. (1946). The people look at radio. Chapel Hill, NC: The University of North Carolina Press.
Lazarsfeld, P. F., & Kendall, P L. (1948). Radio listening in America. New York: Prentice-Hall.
Lazarsfeld, P. F., & Stanton, F. N. (Eds.). (1942-43). Radio Research. New York: Duell, Sloan, and Pearce.
Love, W. (1994, October 28). Balancing act: Breaking new music. Radio and Records, 21, 26.
MacDougald, D. (1941). The popular music industry. In P. F. Lazarsfeld, & F. N. Stanton (Eds.), Radio Research (pp. 65-109). New York: Duell, Sloan, and Pearce.
Maxwell, C. (1994, August 5). Plays per week: How they've changed the industry. Radio and Records, 21, 21.
Maxwell, C. (1995, January 6). The tide has swung back. Radio and Records, 22, 27.
Negus, K. (1993). Plugging and programming: Pop radio and record promotion in Britain and the United States. Journal of Popular Music and Society, 12, 57-67.
Pavia, B. (1983). The program director's handbook. Blue Ridge Summit, PA: Tab.
Peatman, J. G. (1942-43). Radio and popular music. In P. F. Lazarsfeld, & F. N. Stanton (Eds.), Radio Research (pp. 335-396). New York: Duell, Sloan, and Pearce.
Shallett, M. (1990, January 26). Tracking consumer buying habits. Radio and Records, 17, 44.
Smith, W. (1989). The pied pipers of rock 'n' roll: Radio deejays of the 50s and 60s. Marietta, GA: Longstreet.
Rothenbuhler, E. W. (1985). Programming decision making in popular music radio. Communication Research, 12, 209-232.
Rutowski, R. (1989, July 1). "Safe" radio is destroying the medium. Billboard, 101, 11, 73
Wiebe, G. (1940). The effect of radio plugging on student's opinions of popular songs. Journal of Applied Psychology, 24, 721-727.
Wrzesinski, J. L. (1979). Music programming at contemporary music radio stations. Unpublished master's thesis, Miami University, Oxford, Ohio.
Sample Song Table
|Artist/Song: Jeff Carson, Not On Your Love
|Week: || 1 || 2|| 3|| 4 || 5 || 6 || 7 || 8 || 9 ||10 ||11 || 12 || 13 || 14 || 15|| 16|
|A.R.:||a 25 ||20 ||15 ||10 || 5 || 4 || 2 || 1 || 8 || 18 || ** || ** || ** || ** || ** || ** |
|S.R.:||b 24 ||18 || 11 ||10 || 5 || 5 || 5 || 6 || 8 || 6 || 5 || 5 || 5 || 6 || 7 || 8|
|** Denotes the song is no longer on the chart|
aA.R. = Airplay Rank (Airplay Chart)
bS.R. = Sales Rank (Country Singles Sales Chart)