Periodic Reporting for period 1 - ITDS (Intelligent Theatrical Distribution System)
Période du rapport: 2019-05-01 au 2019-08-31
The objective of this project is to draw the missing link between the theaters and the movie makers by shunting distribution companies. With ITDS, Movieday targets the long tail of undistributed movies and acts like the best partner for the movie makers, and their sole distributor. New movie makers are currently rejected by bigger distribution companies because they see a minor return on investment, mostly because their crowd is not apparent – movie makers do not always have a Twitter account or a Facebook with lots of followers to display their popularity. Movie makers are facing difficulties to convince distribution companies to find a screen for them. ITD-S is a platform embedded on our website Movieday.it that enables theaters to schedule movie projections, from the films we already have in our portfolio.
The objective of this Feasibility study was on the commercial side to 1) find the movie makers where they are to increase our portfolio, 2) find the theaters where they are and a way to contact them in the most efficient way to augment the number of screens where our portfolio of movies can be projected 3) and find a revenue model that would best suit both movie makers and theaters. On the technical side, we had to 1) put fan data together in a data lake 2) set the base of the architecture of our machine learning engine so that the event scheduled by the theatre best fits the taste of the local crowd, resulting in an increase of the revenues for the theaters.
To make the engine learn faster, we need to multiply the sources of relevant user data in the theaters’ area to be able to train our machine learning engine and make it learn the final users’ patterns. For this purpose, we plan on developing a ticketing tool to have a customer database on our servers. We will also link ITDS to Facebook’s profile to benefit from the synergies it offers, in terms of users’ movie tastes, movie ratings.
We have already asked 5 theaters in Italy to ask movie goers to fill a form we will carefully design to collect more data about them, their tastes. At Movieday, we are respectful of our movie fans’ privacy, so the information will be anonymized, and the whole data collection will be GDPR compliant.
Movieday's main goal is to distribute the independent movies that can fill empty theaters through the new digital and intelligent technologies that can profile and aggregate the right audience at the right time at the right place, with a high social and commercial impact.
The objective of this Feasibility study was on the commercial side to 1) find the movie makers where they are to increase our portfolio, 2) find the theaters where they are and a way to contact them in the most efficient way to augment the number of screens where our portfolio of movies can be projected 3) and find a revenue model that would best suit both movie makers and theaters. On the technical side, we had to 1) put fan data together in a data lake 2) set the base of the architecture of our machine learning engine so that the event scheduled by the theatre best fits the taste of the local crowd, resulting in an increase of the revenues for the theaters.
To make the engine learn faster, we need to multiply the sources of relevant user data in the theaters’ area to be able to train our machine learning engine and make it learn the final users’ patterns. For this purpose, we plan on developing a ticketing tool to have a customer database on our servers. We will also link ITDS to Facebook’s profile to benefit from the synergies it offers, in terms of users’ movie tastes, movie ratings.
We have already asked 5 theaters in Italy to ask movie goers to fill a form we will carefully design to collect more data about them, their tastes. At Movieday, we are respectful of our movie fans’ privacy, so the information will be anonymized, and the whole data collection will be GDPR compliant.
Movieday's main goal is to distribute the independent movies that can fill empty theaters through the new digital and intelligent technologies that can profile and aggregate the right audience at the right time at the right place, with a high social and commercial impact.
We’ve always known there’s an opportunity with the long tail of movies undistributed.
The feasibility study gave us the opportunity to do a thorough market analysis: we looked beyond Italy, and found out that Germany, France, Spain and the Netherlands are countries we need to target for our European expansion. We’ve identified there’s a huge opportunity for us to be seized, as the theaters are not filling their rooms at each sequence.
Their market potential is huge (TAM = 7bn €), and their box office totaled 3.9bn€ in 2017, meaning that there’s a huge opportunity for us to grab (a percentage of the remaining seats unfilled approx.). We’ve also found out that the rich cultural diversity in these countries can be leveraged to bring people together behind a screen.
We are optimistic we can grab 1% of this market share in 4 years after commercialisation. For this, we worked on the 3 main pillars:
1) The commercialisation plan: we will target associations gathering theaters and movie makers to scale our portfolio. Our objective is to work with 1680 theaters in 2023.
2) the revenue model – based on assumptions, we cross-checked what was the most appropriate business model. It will be a revenue split between the movie maker (20%), the theatre (40%), and Movieday (30%), detailed on the revenue model sub-section of the feasibility study.
Regarding the architecture for the machine learning, we’ve identified 3 stacks we need to work on to improve our proof of concept: the data ingestion pipeline, to acquire all the data on which the prediction algorithms will be based. A part of the data is already available in Movieday (such as the profile of the spectators, the box office of the films, the topics covered, the performances of the cinemas, their correlation with customers’ tastes), but others (such as weather information, other theaters’ data, other people than current customers’ tastes, demographic data) must be acquired from external sources and all the data will then be entered in a data lake from which the prediction engine can carry out its analyses.
The prediction engine, it will be necessary to experiment and implement data prediction / classification algorithms to assess how high the participation in one or more events (or an average of them) will be in a given period and place, in order to evaluate the best programming. The third is the dashboard, a dashboard on which to see the prediction result.
The feasibility study gave us the opportunity to do a thorough market analysis: we looked beyond Italy, and found out that Germany, France, Spain and the Netherlands are countries we need to target for our European expansion. We’ve identified there’s a huge opportunity for us to be seized, as the theaters are not filling their rooms at each sequence.
Their market potential is huge (TAM = 7bn €), and their box office totaled 3.9bn€ in 2017, meaning that there’s a huge opportunity for us to grab (a percentage of the remaining seats unfilled approx.). We’ve also found out that the rich cultural diversity in these countries can be leveraged to bring people together behind a screen.
We are optimistic we can grab 1% of this market share in 4 years after commercialisation. For this, we worked on the 3 main pillars:
1) The commercialisation plan: we will target associations gathering theaters and movie makers to scale our portfolio. Our objective is to work with 1680 theaters in 2023.
2) the revenue model – based on assumptions, we cross-checked what was the most appropriate business model. It will be a revenue split between the movie maker (20%), the theatre (40%), and Movieday (30%), detailed on the revenue model sub-section of the feasibility study.
Regarding the architecture for the machine learning, we’ve identified 3 stacks we need to work on to improve our proof of concept: the data ingestion pipeline, to acquire all the data on which the prediction algorithms will be based. A part of the data is already available in Movieday (such as the profile of the spectators, the box office of the films, the topics covered, the performances of the cinemas, their correlation with customers’ tastes), but others (such as weather information, other theaters’ data, other people than current customers’ tastes, demographic data) must be acquired from external sources and all the data will then be entered in a data lake from which the prediction engine can carry out its analyses.
The prediction engine, it will be necessary to experiment and implement data prediction / classification algorithms to assess how high the participation in one or more events (or an average of them) will be in a given period and place, in order to evaluate the best programming. The third is the dashboard, a dashboard on which to see the prediction result.
Movieday is the first company bringing talented movie makers together with theaters, in an intelligent manner. Our machine learning tool already learns from us, employees and managers of Movieday, true movie fans, and we verified an 80% match between our choices and the platform choices. The platform comes at no-cost for the theaters, and we are supporting the promotion of the movie for the movie maker.
By the end of the project, we will be able to provide movie forecasting, optimized booking, film insights, and an optimized scheduling to increase the profitability of the theaters.
The risk theaters take with displaying a movie on a screen is virtually transferred to Movieday, but the platform can tell if the event will be successful or not for the crowd in the surroundings, so the risk is evacuated by the accuracy of the prediction engine.
We will sell a different offer to our direct customers, the theaters: different kind of events, for educational purpose: around social subjects, politics, economy, that would follow the program of the national education ministry in every country.
We would do recreational movies as well, from every genre: drama, horror, action, humor, love movies, short films, coverages from independent journalists.
We’ve found out that the fewer the total population in a city, the lower the cinema frequentation. It means that we can leverage ITDS to make people go back to the cinema in the country side.
The societal implications would be tremendous: it would facilitate social inclusion, promote culture by talking about sensitive subjects, and it would bring communities normally excluded in a foreign country back with the locals.
We plan on working with the French community in Italy, in Spain, with the Chinese community in Spain, and the Italian community in France, and finally with the Turkish community in Germany, to show new movies in our network of movie makers, or classic movies.
By the end of the project, we will be able to provide movie forecasting, optimized booking, film insights, and an optimized scheduling to increase the profitability of the theaters.
The risk theaters take with displaying a movie on a screen is virtually transferred to Movieday, but the platform can tell if the event will be successful or not for the crowd in the surroundings, so the risk is evacuated by the accuracy of the prediction engine.
We will sell a different offer to our direct customers, the theaters: different kind of events, for educational purpose: around social subjects, politics, economy, that would follow the program of the national education ministry in every country.
We would do recreational movies as well, from every genre: drama, horror, action, humor, love movies, short films, coverages from independent journalists.
We’ve found out that the fewer the total population in a city, the lower the cinema frequentation. It means that we can leverage ITDS to make people go back to the cinema in the country side.
The societal implications would be tremendous: it would facilitate social inclusion, promote culture by talking about sensitive subjects, and it would bring communities normally excluded in a foreign country back with the locals.
We plan on working with the French community in Italy, in Spain, with the Chinese community in Spain, and the Italian community in France, and finally with the Turkish community in Germany, to show new movies in our network of movie makers, or classic movies.