Kenyan music consumption trends Part 1

Sauti Sol
Shares

The Kenyan Music Industry, just like most of biashara in Kenya is largely informal. This means that it’s hard to regularly get data about it, and interactions/feelings/impressions of it will vary greatly. While some people exclusively experience and love Kenyan music, others may actually never hear it and couldn’t be bothered to look for it; and everything in between.  However, not being able to see the data doesn’t mean that the data doesn’t exist. To stretch the analogy, you being the only customer at a shop at a particular time doesn’t mean that the shop has only one customer. And therefore, you not liking a song, or being the only one you know who likes that song, doesn’t mean that the song is bad or generally disliked. Music is a form of art and art is subjective.

There was a time when people thought that the world was flat, and that if you walked far enough, eventually you would reach the edge and fall off. Anybody who tried to say otherwise was treated somewhat as a pariah who should be burned at the stake. People have a narrative and they want to stick to it no matter what. I feel as though this is always the case whenever we get into these “Kenyan Music consumption patterns” debates online. We tend to assume that our reality is the only reality and discussions tend to always boils down to “well that’s not how I see it” conundrums. In my opinion, a type of Dunning-Kruger effect. But thank God for science and it’s ability to help in these situations.

So I went out and got my hands on a data set printed out of a monitoring system that’s run by MCSK. It’s called Vericast and was installed by MCSK in early 2015. It monitors about 50 Kenyan Tv and radio stations in real time 24 hours a day. It has the ability to recognize songs and log plays. See this link for more information on what this product can do. This particular print out lists the top 100 most played songs across the 50 stations in Kenya being monitored. It also shows their track names, play counts and performer names. The data I have here was taken in the 130 days between April 1 and August 8 2015. You can find and download this data set for yourself here.

Now I would like to use this data set, to tackle some of the most common hypotheses that we see/hear online, and for this we need to make two key assumptions. First of all, we need to assume that a play is equal to 5 minutes so that we can try to create a level playing field among songs of different lengths. I also suggest that we assume that every station monitored by the system is broadcasting for 24 hours each day. We know that there may be some variations to this rule but we need to level the playing field here too. This therefore means that over 130 days, the total airtime available across 50 stations is 9,360,000 minutes.

monitoring-period

Hypothesis 1: “Kenyans prefer Tanzanian, and Nigerian Music over Kenyan Music”

It’s difficult to prove preference unless this preference results in some measurable action e.g a vote or sales. In other words, we can’t really measure whether Kenya prefers Panga soap over Omo but we can measure their sales performance. So for the data we have, let’s look at song plays as sales and airtime as market share. Now, this hypothesis seems to suggest a direct competition for the same in the Kenyan airspace by songs/artiste from these countries. So what does the data say?

data-of-songs-played

Out the 100 of the most played songs in this period, 32 are Kenyan, 15 Tanzanian, 13 Nigerian, 14 American and the rest split among DRC, Jamaica, Latin America, EU, and Zimbabwe. Kenyan music is also leading in total recorded play counts and commands 32% airtime with Nigeria, USA, and Tanzania following. So if preference has a direct link to sales performance, then this would be the end of the argument. But we know that such a link doesn’t always exist. In fact, somebody said to me that there’s a “general feeling” that this hypothesis is true. So now let’s talk “general feelings”.

When talking about manufactured products these “general feeling” discussions are, usually, easily settled by introducing the ideas of brand value and target market. Because it stands to reason that one can prefer something and exclusively use it, without that thing ever being a bestseller. Further, preference for a product does not necessarily mean preference for the company that produces it, or even the country of origin of that company. In short, you can love the iPhone, hate the iMac, detest Apple but love the US. If you take songs as products, albums as brands and artistes as manufacturing companies, then we can apply the same rules. You can love Rock My world, hate Michael Jackson, not care much for the album Invincible and love the US. And it will all be valid. In fact the country of origin of the manufacturer doesn’t really matter. For example there may be a general bad feeling about products from China but the iPhone, that so many love is manufactured there.

Besides all this, there isn’t a prototypical music consumer or even a prototypical Kenyan consumer as this hypothesis seems to suggest. The market is highly segmented for most products and especially for art. So the best answer that we can objectively give in response to this hypothesis is that it depends on who you ask, where they live and what they listen to. The numbers (plays and airtime), however, don’t seem to support this.

The other thing to note here is the numbers on plays per song from each country above. Zambia (Amarula) is leading here with over 1600 plays per song with Nigeria second with 1500 and Tanzania third with 1200 plays per song. This is an interesting data point. Fewer foreign songs but a lot more plays per song. I imagine this is what give the illusion of preference as stated in this hypothesis. To industry insiders there’s a clearer reason for this, which we discuss a little later in hypothesis 3.

Hypothesis 2: “Kenyan artistes don’t produce as much music as foreign artists”

available-minutes

As we have already discussed, over a period of 130 days, the total number of available minutes are 9,360,000. With a total of 99,511 plays, the Top 100 songs only command 495,444 minutes (assuming our 5 min per play rule). This is approximately 5% of total available airtime. 95% of the airtime is doing other things, the most likely being adverts, presenter links, news and maybe songs from below top 100.

So let’s assume that Radio stations spend 70% of their time talking, doing news, selling ads and doing non-music related things (which is a gross over estimation). This still means that songs outside of the top 100 list command 5 times more airtime than the top 100.

percentage-airtime

Now because the stations being monitored here are situated in Kenya, what are the chances that the songs outside of the top 100 are majority foreign? Yes it’s easy to download songs from the Internet but this would mean that a)radio stations are run by people who engage in piracy and b) said piracy is still easier to do to fill your playlist than to receive music from local acts. It would also mean that there’s a huge consumer demand on radio stations encouraging them to be pirates. Which, in my own experience, is a long shot.

That said, I don’t have any numbers on this, but I would assume that radio stations get a lot more submissions from local acts than they do from foreign acts.

I also think that it’s harder to play foreign music on vernacular stations. Vernacular stations are getting more and more popular and commanding more and more of a following according to Communication Authority of Kenya (CA) numbers. It’s my submission, therefore, that these stations are probably contributing the most to the variety that gives this large 25% of available airtime to songs that are outside top 100 list.

Hypothesis 3: Radio plays what the Kenyan audience wants/requests

There’s chicken and egg question for us to consider here. What come first, the discovery of the content, or the request for content? Can one request something without ever having heard it and/or can this happen simultaneously? That is can you hear a song on YouTube and immediately request to hear it on radio? And if so will the radio station have it cued and ready?

Now, at the very top of the list is a song called Sura Yako by Sauti Sol which was released in 2014. In the period of monitoring it garnered 8,723 plays or 43,615 minutes. This translates to an average of 67.1 plays per day or 335.5 minutes of airtime per day. Across fifty stations this means that this song got 6.71 minutes per station which is equivalent to 0.04% of available airtime.. Assuming the 5min playtime per play, this means each station played this song once a day. This is the number one song during this period that had been marketed for at least 6 months.

song-played

But lets’ also look at this in a different way. Let’s assume that urban stations would have a different playlist from rural and vernacular stations. And let’s assume that out of the 50 monitored stations 15 (approx 30%) would be considered “urban”. So, looking at 15 urban stations only and adjusting our math to fit, it means Sura Yako got 22.36 minutes per day or 4 plays per day per urban station. Which is about 1.55 % of the available airtime on urban stations.

days-monitored

Compare this with a song called Amarula, which was released on April 10 2015 and in the monitoring period garnered 1,601 plays or 8,005 total minutes. This means that this song was played an average of 13 times a day, receiving 65min a day. Across 50 stations this is a little over 1min a day per station but across 15 urban stations is 4 minutes a day and 1 play daily per station. In other words it achieved one quarter of the daily airtime on urban radio stations that Sura Yako achieved, despite being released almost a year later and 10 days after start of monitoring period so no benefit of marketing. It also means that it could have well been played 13 times on the very day that it was released.

Another song called Nerea has similar performance. It is released on the 20th April 2014 and received 1,539 plays or 7695 minutes airtime. Averaging about the same 1 play per urban station per a day and receiving 13 plays even on the day that it was released.

If radio stations only play what people request, how would this be possible? It would mean that people are either a) aware of products that don’t exits yet on air and make request ahead of time or b) radio producers have a direct line to all studios here and abroad and have set aside airtime for new songs to be slotted in. Either theory is implausible.

It’s more likely that when it comes to new content, radio stations play a) what they want, b) what they think their audience will enjoy c) what the producer at work likes d) what the producer at work may have been otherwise encouraged to play. Google payola.

Hypothesis 4: “It cannot be a hit if I have never heard it”

Sitting at number 12 in the top 100 is a song called Takomugaksei by Mum Cherop. It was released in 2013 and during the monitored period got about 1157 plays or 5,785 minutes.  Now it’s important to note that because this song is in Kalenjin language, it’s unlikely that it is played say on Coro FM or any urban radios. So for this we would use just the 3 Kalenjin Language stations that I know Sayare radio, Chamgei FM and Kass Fm. Which means that Takomugaskei gets at least 44 mins per day on these 3 stations and equivalent to 1% of the available airtime.

Compare this performance with that of the number one song in that period. So is Takomugaksei a hit or not?

sura-yako

SHARE
Previous articleAhmed Rady is the new Coca-Cola GM for East Africa
Next articleWith EazzyApp Equity Bank continues to disrupt the banking sector
Dan Aceda is a Kenyan singer and songwriter known for his penchant for sweet melody and unique storytelling. He has produced three studio albums and has played at concerts all over the world including the US and Europe, East Africa, Malawi and more. He is the Founder and CEO of The African Bonfire, a multi media production company based in Nairobi Kenya. He is also a current member of the prestigious Global Health Corps Fellowship Class of 2013. As of June 2014 Dan is also a member of the UN Global Accelerator network of entrepreneurs.

1 COMMENT

Comments are closed.