Twitter as service monitor

How do you monitor broadcast output on FM, LW (MW in some areas), two DAB frequencies, LW and FM channels for digital TV on Virgin cable, Sky satellite, Freeview and the online versions of the LW and FM streams in Flash and Windows Media? I’m not a radio engineer so jump in and correct me here if I’ve got this wrong but checking playout for silence is easy enough (and explains why you don’t hear 4’33” on the radio very often) but checking that a particular stream is carrying the right audio is an entirely different challenge. Putting a system into the broadcast chain that could automatically check that all those streams (the list at the top is just for Radio 4) were carrying the right audio at the right time and then take action if it finds an error sounds like the definition of a tall order.

And the answer, of course, is obvious. You ask the listeners. In fact, you don’t even have to ask them. As part of my impossibly glamorous day job, I look after Twitter and Facebook accounts for Radio 4 and at about 0645 yesterday morning, while eating my cornflakes, I noticed a lot of tweets mentioning Radio 4 and complaining that there was something wrong with the network’s FM stream on Internet radios and mobiles. These listeners were hearing classical music when they were expecting Today. Apart from the obvious annoyance, some were evidently freaked out by the presence of sombre music (Brahms was mentioned) on the BBC’s primary news outlet – they feared a national tragedy! I responded to some of these tweets, getting a bit more information from affected listeners, then emailed various people I knew to have responsibility for output at BBC radio.

I also tweeted on the Radio 4 account, keeping followers up to date with my actions. There was a to-and-fro of tweets from listeners (using semi-private @replies and public @mentions as appropriate). Although there was no official line on the problem yet, between us we were able to figure out that the only streams affected were the Windows Media ones used by devices that don’t support Flash. So we got a sense of the size of the problem too (pretty small). And somewhere in the middle of all this, BBC staffer James Hart (@syzygy on Twitter) noticed all the fuss and picked up the phone to exactly the right engineer who proceeded to switch the streams back to their proper locations. By 0915 everything was back to normal.

So, to summarise, in this particular minor crisis, Twitter (and Facebook to an extent) did the following:

• Alerted us us to the existence of a problem within minutes of it arising, outside working hours.
• Provided some quite specific data on the scale of the problem, platforms affected etc.
• Gave us a real-time, two-way connection to affected listeners.
• Allowed us to update other listeners about the problem quickly.
• Got the problem fixed quickly.

Picture by Paul L. McCord. Used under licence.