Drive traffic to your platform and boost subscriptions
Leverage Filmaster API · Enrich existing EPG · Real time capability · Open new marketing channes
How does Filmaster work?
Filmaster employs a hybrid of algorithm types that drive our recommendation engine. We analyze viewer preferences, group-trends, and inter-user comparisons in order to predict the best possible movies and shows that a viewer would want to watch. Filmaster data is crowdsourced, and provides the metadata and logic matrices required to power the engine. Our service is based in the cloud, meaning our customers benefit from simple integration, immediate access, and a powerful sales tool that can be deployed anywhere, at any time, on any platform.
Our content-based algorithms deliver recommendations based on a vast array of movie and show tags (e.g. cast, director, genre, year, etc.) that a viewer has previously enjoyed. The algorithm sets values on the importance of characteristics to the specific user and generates a list of predictions.
Those predictions are used to create a list of personalized recommendations for each user, who can then manually narrow results by their mood or other filtering criteria. As users continue to watch content, the algorithm constantly learns to evolve recommendations.
The collaborative approach attempts to narrow specific predictions for an individual viewer by analyzing data gathered from the crowd. Filmaster’s crowd-sourced algorithms and data are the result of 3 years of content analysis, expert curation, and a huge database of ratings which cable providers can utilize immediately upon implementation.
This solution eliminates the expensive and time consuming R&D costs of building recommendation platforms in-house.
Data in the cloud
We’ve been crowdsourcing and curating film data for over three years. Our data sits in the cloud which means it’s immediately available and accessible at all times to our customers. Most recommender solutions will only provide the API without data. Without proper data, recommendations are far less powerful.
Proper data gathering done in-house could take months of R&D in order to accumulate the amount of data needed to provide meaningful results.
Ensure Quality Recommendations
The curative approach reduces the risk of false-positives that can occur (e.g. when an erroneous tag is ascribed to a movie, or a false relationship is made between two movies). Curation also provides a human-layer to algorithmic learning by providing additional details and innovative relationships that computer AI cannot determine alone.
Learn from viewer behavior
The Filmaster engine is capable of enhancing recommendations even further by analyzing viewer behavior. Implicit data about a viewer’s behavior (provided at the discretion of the customer) can be used to personalize recommendations even further. A few examples of meaningful viewing behavior include: How long a viewer has watched a movie or show; What volume the viewer prefers; What time of the day the user watches television.
No cow paths
We’ve made sure that our platform encourages content discovery in order to direct users to new and interesting content. Viewers should never circle the same recommendation category over and over. Note that Filmaster provides various levels of configuration that allows the customer to decide how aggressive or passive their discovery engine will be.
Leverage Filmaster Technology
Convert Non-viewers into Viewers
80% of television viewers turn on their TVs without specific intent for what they want to watch. The term “channel-surfing” is a result of the fact that viewers are most often unaware of the content available to them. Simply put, viewers enjoy watching television, not surfing television. Filmaster provides modern user experience solutions to best navigate viewers around their personalized content:
- Use Filmaster recommendations to boost or enrich your existing EPG without an interface redesign.
- Filmaster channel-based recommendations provide real-time recommendations to your users immediately. Direct users to content from live TV, VOD, and external sources from your service.
- Filmaster is able to develop a recommendation interface where users can access their recommendations, favorites, etc. at any time.
Make an Impact on Subscribers
Filmaster recommendations are powerful sales tools that enable you to leverage existing sales channels and open new ones. Our recommendations will boost your subscription numbers and provide the analytics necessary to maximize the impact of your advertising campaigns.
- Use Filmaster recommendations to increase sales of channel bundles and packages by personalizing advertisements for each viewer.
- Use Filmaster recommendations off-screen in your newsletters, paper bills, and electronic correspondence to drive targeted advertisements.
- Use Filmaster recommendations to improve your current offers to existing customers by showing that your service is always personalized to their tastes.
Data is a Part of the Deal
Data is a part of Filmaster’s offer and is as important as the algorithms. It allows for instant recommendations (as a solution to the “cold start” problem) and unique features such as mood-based recommendations and innovative ways to present content. The data provided by Filmaster—movie and shows traits, information about viewers’ tastes, and interconnections between movies and shows—allows you to compute the personalized recommendations even without any initial data delivered by the customer.
Out of the Box and Running
The Filmaster SaaS model means the product is not only out of the box but actually is ready to use from day one. The customer takes control over different configurations (e.g. whether the algorithm should be more brave or more conservative). Also, the algorithm is modular, meaning its different parts can play more or less signiﬁcant role when used by different customers delivering the best user experience for their consumers.
Simple Pricing Plans
The pricing is ﬂexible and depends on the API usage. It’s important to realize that the SaaS price covers both hosting and support, so it’s very competitive compared to in-house development or using out-of-the-box products which still need to be maintained in-house.
|Ratings/Activities||20 k||10 mln||20 mln||40 mln|
|Generated Recommendations||10 k||4 mln||8 mln||16 mln|
|Users||1000||100 k||300 k||1 mln|
|Guaranteed peak performance (per second)|