VARIETY... IS IT ALWAYS THE SPICE OF LIFE? - Part 2 Greg Diamond In Part 1 we made the assertion that variety in its purest form - a greater number of disparate selections - is actually detrimental to a station’s overall strength. Whether you believe that or not, let’s examine the matter further and apply mathematics to provide more insight.
Taking a page out of your sales staff's book, we’ll examine a song as they would apply a reach and frequency analysis - they are both pieces of audio, after all. It should be stressed, however, that a listener tunes to a station for music, not necessarily commercials. Therefore, while the mathematical premise is still applicable, the manner in which the data is accepted and applied must be altered accordingly. Nevertheless, if you feel someone may hear a song too many times each week, then the following may come as a surprise.
Any demo can be used to calculate the desired plays. You may wish to zero in on your core to ensure they are being super-served. By using MicroBBM or some other ratings software, we can look at a station’s P1, P2, P3, or combinations thereof to determine what percentage will hear a song once, twice, or three-plus times. For our calculation we will also assume your music scheduling software is set up properly to achieve the appropriate daypart "bounce" for the desired category and thus achieve a lower fatigue factor. Finally, we will also assume you are playing the strongest record(s) for your station - a key ingredient under any circumstance and one which your ByrnesMedia consultant will spend many hours helping you to achieve.
By highlighting our imaginary station’s three heaviest-usage quintiles (5 [Heavy], 4 [Medium-Heavy], and 3 [Medium]) and combining them for accurate representation, we can begin the process of examination. Quintiles, as the above parenthetical addition indicates, correspond in opposite fashion to what we commonly refer to as P1, P2, and P3. Quintile 5’s are P1’s, quintile 4’s are P2’s, etc.
We’ll start by looking at how many times we can conceivably play our heaviest rotating category each week.
By selecting 10 occasions (plays per week), we see that 20,800 listeners out of 35,700 (58%) heard our song once. A further 8,100 (23%) heard it twice and a total of 6,800 (19%) were exposed to the selection 3+ times. The average frequency was 1.8. Now, as a programmer, you need to ask yourself if this is sufficient for your needs, since different formats require different levels of exposure. Regardless, we can see that 10 plays per week does not make as much impact as we may suspect.
Let’s take the same group and try for an average frequency of 3. This is a simple case of increasing the occasions until the desired number is reached. In this case, we need 26 spins for an average listener to have heard the record 3 times. Examining the numbers closer, we see that 37% of the audience heard the song once, 23% heard it twice and 40% heard it on 3+ occasions.
It should be noted that the above calculations were performed for a 6am to midnight time block only. Adding another 25% to the number of plays to account for the overnight period gives us a total of 32.5 plays per week. Coincidentally, using rock stations as an example, it is common to run a five song heavy category loaded once into every hour. That works out to just over 33 plays per week.
Now, let’s take it one step further and recalculate using only the P1 group.
In this instance we see that it only takes 17.5 plays to achieve an average frequency of 3 with 36% having heard the song once, 19% twice, and 45% 3+ times. Keep in mind, though, that these are your heaviest users and as such a higher repetition tolerance would not only be expected, but likely desired.
If we once again combine our 3, 4, and 5 quintiles and then take the same approach to our recurrents we get some interesting results.
For a recurrent that airs once per day we find that 73% of the listenership heard this very important element of the station’s music run once. A further 19% heard it twice and only 8% heard it 3+ times. The average number of exposures was 1.4. Recurrents bridge the gap between the old and new components of a library. They have the freshness of a current and the near-familiarity of gold, so playing a recurrent once a day is quite safe... possibly even too conservative.
We can apply this principle to gold as well but I think it’s clear we can sometimes be overly concerned with music spin frequency. Remember that station staff, who often opine we are playing a particular song too much, are not your average listener. They are tuned to house monitors and get a much higher than average exposure to your playlist.
These calculations should help to show how music libraries need not be overly large to achieve the desired "variety." To weigh down a universe with peripheral records only makes it harder to get the "important" songs exposed. As it appears from our calculations, it is difficult enough!
Next month we will examine how the common, "You play the same song over and over!" complaint has contributed to the bloated universe and how such complaints should be viewed. Also, in upcoming articles we will give you tips to better utilize your existing universe to gain "perceived variety", instead of just adding more records. And let’s not forget about the Jack/Bob/Joe/Pete/Phil/Walter/Sydney format. Why does its "variety" work?
Happy Holidays everyone and here’s to a successful 2006! |