Playing the Music You Want to Hear

I once heard someone say that there are two things in volunteer radio stations that can cause big arguments – jingles and music. While I do believe that’s somewhat unfair, experience tells me music choices in a station will be regularly questioned.
It’s not often I write stuff from a programming point of view, but I’m going to here. It’s a whole bunch of stuff I’ve learned over the years, a number of them wearing the volunteer PC hat. Some of it you might not agree with and other bits will sound like common sense.
That said, it’s the approach we take programming this station.
With my asbestos suit donned, we better start at the beginning. Who is the station trying to entertain?
With NHR, the remit is simple – interact with and entertain the patients of Nottingham’s QMC and City hospitals. Regularly pounding the wards really helps with this but what about when we’re in automation. That’s the majority of hours we can’t get volunteers in.
We could fill with hours of material we’ve accumulated over the years. However, a lot of it is dated or can be less engaging than back to back music. As a couple of examples, we had an in-house “A to Z of Pop” that ran five minute links between songs and only got to the letter C before the volunteer gave up.
We also had a show that consisted of regurgitating the Guinness Book of Hit Singles, a book I’ve considered banning from studios in the past. In the case of this presenter, I made him listen back to what he’d done on air (all part of the normal training/feedback we do). It was so boring he protested listening to himself and finally dawned what was wrong with the show.
I got a lot of flack when I took over and made some rather sweeping changes on air. However, examples like this remind me of where we came from. It’s been a lot of work from the team and even resulted in the odd HBA award along the way.
Back to the music and we put a lot of effort into what we do at NHR. After all, we’re in automated music almost as long as stations like Jack and SAM FM.
It may sound nuts but it’s working. Combined with actually going out on the wards, we’re seeing a vast increase in hours and reach. Step onto some wards and we’re the background noise chosen by staff and patients. We’re also very quick in getting phone calls when we go silent. I’d call that relative success. After all, people are always much quicker to complain than celebrate.
When it comes to programming music we start with demographics. I know, I hate the term as well. However, it’s pointless playing back-to-back current hits to an older audience.
A scary number of years ago, when I first stepped into hospital radio, I was told that our listeners were all 50+ with an average age of 65.
How accurate is this? Well, I did do some number crunching a few years back using publicly published figures from our local NHS trust. At the time, the average age of someone on a long term stay in hospital is 54. Over 65s were twice as likely to be staying long term but the numbers drop off quickly for obvious and depressing reasons.
The result of all this – targeting a 50+ audience music wise isn’t such a bad idea after all. That means you’ll want to be classic hits heavy.
If you do the sums on someone aged 65, they were born post war (sorry guys, time to drop the 1930s ditties) and generally followed the music of the 60s and 70s.
Admittedly, it’s an average we’re talking about here. You’ll find many older ladies requesting the boy bands of later years. Something, surprisingly, that’s backed by the top requested 50 artists chart we publish annually.
So, we come round to that much dreaded format – 60s to today. The approach NHR takes to this is to be pop/rock heavy. After all, it’s the stuff that’s really been big over all these years.
On that note, I really don’t subscribe to the idea that hospital radio should be focussed on ticking as much boxes as possible. There’s nothing wrong with a specialist music show with a passionate and knowledgable presenter. In fact, I welcome them and they’ve made some of the best radio I’ve heard over the years.
However, specialist music for the sake of it is, quite frankly, nuts. I’ve never seen anyone suddenly get into light jazz because they’re ill. The people we entertain on hospital radio are the same people walking the streets every day. Tastes are therefore likely to line up with the mainstream.
On top of that, familiarity and that connection back to reality are things hospital radio should be trying to deliver. Playing music that’s too far off the beaten track too often isn’t going to do that.
And that’s one of the things we’re bearing in mind when deciding if a song should go on active rotation. In fact, there’s a list of criteria we try to test every song against:

• Research vs. gut instinct. Is this regularly requested? Did it do well on the Top 50 artists list? Either way, we tend to rely more on gut instinct than surveys.
• Is it familiar? We’re back to the issue of do our listeners know the song. Too many misses in a row and we’ve lost them to Ken Bruce.
• Was it a big hit? That doesn’t mean play every number one out there. However, that song that spectacularly missed the Top 40 probably isn’t for us.
• Has it survived the passage of time? Novelty songs tend to fall short here. Some songs are of their time and really don’t sound great now. That’s ignoring issues about artists it’s understandably tasteless to play the music of nowadays.
• Does it fit our station style? Something that sticks out like a sore thumb such as an incredibly heavy rock track or electro-house hit won’t pass muster.
• Is anyone else playing it? I get a bit of grief for this one because “we should be playing songs no-one else is”. However, if no-one else is playing it, is that for a good reason?

Using this criteria will often result in a smaller music selection than you started with. It’ll also lead to the complaint about sounding repetitive and “same-y”.
We operate with around 2000 to 2500 songs on live rotation at any time. With a bit of clever scheduling, our tightest rotation is around two and a half days for big hitters. At the other end of the spectrum, some spice songs appear about once a month.
Having tens of thousands of songs on rotation isn’t necessarily going to make your station sound better. Before I took over, we had no flow between songs and no real filter on what was played. Songs like Goldie Looking Chain’s “Guns Don’t Kill People” made a number of appearances on air.
Either way, once you’ve got your songs and you’re sure they fit the station sound you’re shooting for, the next stage is categorisation. This is one I’ve seen done in a couple of different ways over the years.
The first is one big “music” category. There’s literally no control over what the computer will pick next. It’s the Liquorice Allsorts in a bag of Revels approach to programming music. It works and you get a surprise every time. The catch is that it’s usually a bad surprise – coffee and liquorice never made a good mix.
It’s also a really odd approach to take. I’ve seen stations put tons of effort into live programming and then just hit “shuffle” for the automated hours. I know resources are limited in the voluntary world but a small effort here will have a surprisingly big pay-off.
The second approach I’ve seen is simply to break your music into decades. One group for 60s, another for 70s and so on. While it does allow some control over the sound of the station, and also ensures you get the decade balance you’re after, there is a slight problem with it.
What’s an 80s song? Is it Jane Weidlin – Rush Hour? Is it Chris de Burgh – Lady in Red? Is it Alice Cooper – Poison?
The answer’s all of the above. You might gain some control over the era and age of songs, but you rely on scheduler rules to control flow. Not world ending but something to bear in mind.
A third approach I’ve seen to programming music radio is to program by pace/style/genre. This works best for specialist stations where you want to get the balance right. It’s pretty much the approach I’d look to take on that classic rock station I’d love to be programming.
At NHR, we do things a little differently. Well, if you call combining a couple of the above approaches doing something different.
This is the “secret sauce” in getting the sound we want. We call it the NHR grid:

Along the top there’s the pace/style of the song. Most pop songs fit on a scale of slow/easy/pop/party. Rock songs are a little different and sit out on their own.
Down the left hand side, you’ll see the era. Gold is anything up to the mid 80s. Classic covers the mid 80s to 2000. Modern is anything in this century.
With this arrangement, it’s possible to “ramp” songs of various ages without worrying too much about the exact year. We really don’t care if we get a 60s or 70s song. We care more about getting an “Gold” song.
If we need to play songs from a specific decade, we use sound codes to restrict the song selection. That allows us to maintain the flow and pace control we get from this system. It also allows us to apply the same approach for “Best of British” and similar other shows.
The control we get is surprisingly good. With the pace controls, it’s possible to schedule slower hours (through the night) and faster hours (breakfast) by simply adjusting which categories we pick on a clock. Ensuring we remain classic hits heavy is as simple as picking more gold songs in an hour.
In order to maintain some flow, we have one simple rule for this grid – you can only move one square in any direction. While your move can be diagonally, it prevents any big jumps from occurring.
There is one exception to this rule – the rock category sits out on its own. We still maintain the era rules on the grid but rock songs can be scheduled between pop and easy songs.
The result of all this is a station sound that controlled but unpredictable. You might be able to guess what type of song is going to play but it’s hard to spot any pattern and guess when ABBA o’clock is.
What’s more, we do this with only five general clocks and a small number of “specialist” clocks (e.g. rock night, non-stop decades, Best of British).
It’s also a scheme that works on all sorts of music scheduling tools. I did the original work using Rivendell’s built-in tools. Nowadays we use GSelector but the basics are the same.
That said, we’ve really benefited from the category “level” features. In the past, we “bodged” it with song codes and a couple of rules.
Nowadays, we use these features to play the bigger hits more often. Spice songs don’t sit in a separate category (and thus become predictable). Instead, we make them low rotation in a wider category.
All of this work results in a station that sounds good, flows through every segue and seems to go down well with its audience. It took a lot of work to get there (literally years) but the results make it worthwhile.
If you listened to NHR in automation, I bet we sound better than you’d expect from hospital radio. Even if we are occasionally compared (musically) to this lot.