The news that the BBC is looking to renew sketch show Horne & Corden has been met with many raised eyebrows. While the opening show brought in record numbers for a first episode of a comedy series (817,000), it lost almost half the viewers over the six episode run. Coupled with a poor critical response, this did not look like a show that was ready to return.
While the BBC should not be chasing ratings for every show, there comes a point when any switched-on organisation has to look in the mirror and say, “we tried our best, it wasn’t enough, move on,” and let a program die a natural death.
So how has the BBC come to decisions like this one?
Alongside the ratings, the BBC also has an Audience Appreciation Index (AI). This samples a few thousand viewers and eventually gives each programme a score from 0-100 relating to how much they liked the programme. A score of 70 is about average.
Originally provided by the Broadcasters Audience Research Board (BARB), each programme’s AI score is now researched in-house by the BBC. It has become a shadowy, hard-to-find number which carries an almost mythic quantity for fans of programmes with poor ratings. If the audience love it so much, then a Corporation which does not chase ratings should rely on that for commissioning.
This is still a coarse measurement, not just because of the small sample size, but the single AI number cannot be broken down into why the audience appreciated a programme. Did they like it because it was funny? Entertaining? Made them think? This problem has been there for many years — witness the Annan Committee’s comments in 1974 that “[working] on audience appreciation is too broad a sweep.”
That has never stopped the BBC making a big deal of AI numbers when the score into the high eighties, such as the recent Doctor Who Easter Special and its AI rating of 88.
So are discussions on a second series of Horne & Corden based on the strength of a strong AI number? We don’t know. The press release for the end of the series from the BBC talks about “an average of 0.9 million viewers” and “the most successful ever first series of comedy on BBC3.” Unlike the good Doctor, there are no AI numbers easily available.
It is notable that the Corporation have added the ratings for all the airings of each episode together to come up with that average audience. While it’s impossible to be arithmetically precise to judge a success, the numbers give the appearance of a Sir Humphrey looking to save a favourite programme no matter what.
The BBC does have a public service remit in making programmes for smaller audiences, but it would be nice to see some transparency in the decisions, especially those where the stars are on large contracts. The transparency and independence of the television ratings is maintained by BARB, but the AI ratings and other internal data used by the BBC remain just that — internal. When the numbers, be they ratings or AI, are good, out they come. When they don’t match an argument, then something else needs to be found.
Do the numbers lead the decision inside Broadcasting House, or are numbers chosen that support a certain viewpoint. Or are commissioners relying on… courageous decisions?