Our film campaign tracking product Cinematix is now two and a half years old. What started as a five-city weekly track has evolved over time into a 19-city daily track. It has been a satisfying experience to see Cinematix grow and be adopted by the industry as its campaign tracking currency, with a strong box office forecasting ability.
This week, we took the plunge and added a new parameter to Cinematix, called FBO, standing for First-Day Box Office. FBO will forecast the nett domestic box office collections of forthcoming films, upto seven weeks in advance.
A few questions related to FBO, whose answers you may want to know:
We have been testing and evolving the FBO model over two years now. Subscribers of Cinematix, as well as industry folk in general, have often asked us to forecast or estimate the opening day or weekend collections of films. We have shared the forecasts on request, sometimes even tweeted our forecast. However, it’s only now that we believe that our model has passed the relevant statistical tests and hence can be used for regular forecast for all films.
What is the model based on?
FBO is based on a set of inputs, led by the Cinematix scores of Buzz, Reach, Appeal and Interest. It also accounts for competition films, scale of release, audience profile, festival release, non-Friday release etc. A complex interplay of parameters is what FBO is all about.
Why only first day, not the weekend?
All inputs in FBO are “marketing” inputs, and are independent of the content of the film. However, we have seen that the content begins to play a role very early in a film’s box office performance. A film can gain upto 20% on Saturday and upto 35% on Sunday purely on strength on its content. The first day is the only day where the role of the content is minimal, to the extent of less than 5%, and hence, campaign parameters are enough to forecast box office accurately.
While we test a lot of films and hence have a measure of their content strength (Word Of Mouth), that data is available only to the producers of the film being tested and is not a part of Cinematix. In Cinematix, the WOM data is collected over one week after release, and is hence not an input into the FBO forecast.
Will FBO be available for all films?
We will refrain from reporting FBO for films that are likely to open at less than Rs. 50 lac on the opening day. At low collections, error margins tend to be high, and hence, should not be reported. For all other films, FBO will be available everyday from Wk -7 stage (seven weeks before release) or within three days of the start of the campaign, whichever is later.
Will it always be accurate?
FBO is based on a statistical model. Definitionally, all such models are mathematical representation of reality. Hence, there will be inherent error margins built into any such model. FBO has been tested to forecast the first-day box office of 85% films with less than 10% error. So, we should get more than 5 out of 6 films right, within 10%. Of course, there are some which we will get right with less than 1-2% error. But there will be that odd film that will be an outlier to the model. All models have an aspect of self-learning built into them. And FBO too will get more accurate with time. We will keep the error margins on FBO transparent all the time.
Will FBO forecasts be tweeted?
The million dollar question! The answer, unfortunately, is “rarely”. Cinematix is a paid service and only limited data can be made public to non-subscribers. Selected FBO scores may be tweeted with permission of the subscribing studio/ producer.
For any more questions on FBO, you can leave a comment here or get in touch on our twitter handle @OrmaxMedia