Randomizing advertisements

This 7.5 minute break in the middle of an IPL innings is a bad idea. The biggest problem is that everyone knows the exact length of the break, and can use it to do stuff – like cook, or clean, or crap, or fag, or maybe watch the Everton-Man U shootout. 7.5 minutes is a lot of ad time, but the problem is that absolutely no one will be watching them. So if you were a smart advertiser, you wouldn’t want to put your ad in that slot – you are better off taking an over break slot.

Now what I propose here is not applicable to cricket – at least I hope it’s not since conventionally you can’t slot ads whenever you want to (Lalit Modi thinks he can change that, though). I don’t know if this concept has already been implemented, and I’d be rather surprised if it hasn’t been. The basic idea is to randomize the length of advertising slots.

So you are watching your favourite soap and there’s a commercial break. And you go off into the kitchen to make a cup of tea. But you don’t really want to miss even a minute of the action, so you’ll go only if you know that the advertisements will go on for two minutes. Historical data tells you that the ads will last for two and a half minutes, and off you go. Now what if suddenly tomorrow there is only twenty seconds of advertisements and you end up missing a bit of the action? You curse yourself, and the soap, and the TV channel, and the TV, and Tata Sky, but you make a mental note not to go make tea during this break the next day.

Now, by randomizing the length of advertising breaks, channels can ensure that people actually watch the ads. If you don’t know if the break will last twenty seconds or two minutes, you are likely to sit glued to the TV, watching the same channel dishing out the ads. You are unlikely to go off to make tea, or to crap, or to channel surf, if you don’t know when programming might start next. You occasionally get pained – when the breaks are too long – but on the whole you end up watching most of the ads.

Yes, there is the chance that the viewer gets pained when the random length for ads that gets picked turns out to be really large. Also, if we shorten a few ad breaks, we should also lengthen a few others? Or increase the number of ad slots? Not really – is my argument.

The clincher here is that by randomizing length of ad breaks, you are increasing the TRPs for the ads! Yes your program may have high TRP but does that normally translate to ads? With this randomization procedure it does. And when this gets established, you can start charging higher for these slots. And if on an average you can charge a higher rate per second of advertisement, you can sure continue to run the program with a smaller number of ads?

It’s win all around. Customer wins because he gets more programming time than ad time. Advertiser wins because he gets more eyeballs for his ad. TV channel doesn’t lose since the loss of revenue from lesser number of ads is more than made up by the higher rate charged on the ads. In fact, by “holding” the customer, the channel ensures he continues watching this program rather than go off on a tangent while channel surfing.

Normally, I try to show situations where everyone can win by reducing the randomness in the system. This case is opposite. By introducing randomness in the system, everyone wins! I wonder if there is a fallacy here. Or maybe what I’ve written here is so obvious that everyone is implementing it and I’ve failed to notice since the only TV I see is sport (not american sport) which has fixed ad breaks.