We’re all familiar with the banality of evil (shout out to my Hannah Arendt crew!), but less well known perhaps is the ‘evil of banality’. An example of this might be starting a blog with the observation that Christmas seems to come earlier every year, and calling it: “It’s Beginning To Sound A Lot Like Christmas”.
Thankfully for me, for weak puns, and for the sanity of listeners everywhere, Christmas actually seems to be arriving later, musically speaking, and apparently we have streaming to thank for that – or data anyway.
Over the years I’ve managed Christmas music programming for radio, TV and streaming, and it’s fascinating to note how the differing levels of listening data for each lead to strikingly variant levels of sanity with respect to when Christmas actually begins.
My tenure as Head of Music at MTV spanned three holidays. The custom, when I took over the music programming team in 2006, was to start playing Christmas videos around mid-November. Another fun, seasonal tradition was an unspoken battle of wills with our main competitor, Bauer TV (owners of Magic, Smash Hits, 4Music, Kiss TV and others), as to who would cave first and go full-blown festive across the board.
That first year, Bauer went a week early and trounced us in the ratings. I had come from Radio 1, where playing Christmas songs before December was considered a cardinal sin, but I soon learned my TV lesson and brought Christmas forward yet another week the following year, turning the tables on Bauer and enjoying a week of seasonally boosted BARB.
In the third year, going yet another week early would have meant playing Christmas videos in late October. We considered it (we really did), but in the end I put my foot down, insisting that playing ‘Last Christmas’ before Halloween was getting a tiny bit ridiculous.
That was a mistake. Bauer went before us – yes, in the last week of October – and their ratings went through the roof. If you’ve ever wondered why Christmas music seems to creep forward every year, it’s because it works.
Another name for BARB ratings in TV-world is ‘overnights’, so called because they are measured electronically and updated overnight. Coming from radio, which in the UK measures itself using the quarterly diary method, the frequency and accuracy of overnight ratings were a blessing and a curse.
A blessing because finding out each morning how the previous day’s programming performed meant being able to make schedule adjustments on the fly. VH1 Classic down this week? No problem! More power ballads. A curse because knee-jerk reactions were common. New host not performing? Get rid.
UK radio, by contrast, finds out how its Christmas programming went down some time the following spring, which is to say it never really finds out at all – things have moved on by then. Faced with a comparative lack of data, sanity seemed to prevail somewhat.
At Radio 1 we would rotate a smattering of Christmas music in early December and then gradually ramp things up over the course of the month until Christmas Day, which would be almost entirely seasonal music. On Boxing Day, Christmas died.
Things are different in the US, where radio ‘flips the switch’ much earlier as a rule and where, interestingly, PPM measurement more closely resembles BARB. This excellent Five Thirty Eight article and the graph to the right show just how crazy things are getting there too.
So what about streaming? Well, streaming services have one challenge that programmers for radio and TV do not – making sure people don’t hear Christmas music for about 11 months of the year. You don’t need me to tell you how loathsome it is hearing Jingle Bells piped into a department store in October; hearing it on an algorithmically programmed radio station in March is even worse.
But with respect to the start of Christmas, does even more data – the deep, granular, real-time listening analytics available to streaming services – lead to even higher levels of seasonal insanity or less? Mostly less.
At Last.fm we did a great data visualisation each year called ‘Is it Christmas yet?’ that calculated the level of actual Christmas music listening as a factor of its peak volume. (Remember Last.fm is scrobbling data from all digital music services.) Each day you could check in and see that it was 20, 30, 70 per cent Christmas, like a kind of data-driven advent calendar.
‘Is it Christmas yet?’ grew out of on a 2011 study showing that, musically speaking, Christmas seemed to be getting later and later each year. ‘50 per cent Christmas’, calculated as a 7-day rolling average of listening to the Top 10 seasonal songs and expressed as a percentage of their Christmas Eve peak, seemed to move from November 30 in 2005 to December 6 in 2011:
More recent data from Spotify For Brands – I’ve grabbed a section of their data vis. below – suggests likewise, with Christmas listening hitting ‘consistent heights’ around November 28 (for some reason the time-based graphs in their data visualisation seem to run right to left), and that ‘Last Christmas’ by Wham! is the most listened-to song during that period. We don’t, of course, know the extent to which this is affected by Christmas programming appearing in Browse.
At the time of writing (November 11), none of the major streaming services are featuring Christmas playlists in their frontline curation – nothing on Spotify, Deezer, Rdio or Apple Music, though of course Christmas music is available on-demand all year round. Last.fm, though something of a footnote in the space these days, is probably still the best reflection of actual demand for Christmas being almost completely devoid of editorial. (Not for lack of trying by me, it must be said.)
TV pushes ahead, albeit returning to somewhat saner waters than in 2008. According to its schedule, MTV is counting down to a pop-up channel called MTV Xmas, launching on November 14 and featuring wall-to-wall Christmas videos. MTV’s current head of music tells me that when it premiered in 2013, MTV Xmas delivered the channel’s highest December ratings in eight years.
All of which is a very roundabout way of saying that no data at all (as for radio) might actually be preferable to some (as for TV) and on a par with loads (as for streaming). I will say this though: linear radio doesn’t need a Spotify data vis. to discover that ‘Last Christmas’ by Wham! is the most popular Christmas song. It has known this for years.