Gamechanger: Personalized Video Content
Pipeline recently had a chance to have a conversation with ContentWise, a global provider of digital content personalization solutions. ContentWise is a spin-off of Moviri, a global
provider of data mining and optimization solutions for enterprise systems and digital media. This unique combination of experience positions ContentWise at the edge of the personal content
revolution. It also places them directly in a torrent of industry trends such as big data & analytics, customer experience management (CEM), and video & content. It's been said that luck is
when opportunity meets preparation and, if nothing else, ContentWise was prepared when the opportunity presented itself.Pancrazio Auteri is ContentWise's CTO. His background places him at the forefront of internet technology and TV services, including a tenure at TiVo where he drove product management for its VOD, metadata, TV Everywhere and third-party content applications.
Pipeline: How did ContentWise come to be?
Auteri: We were helping telecom service providers and other large data center operators predict how resources are going to be used so they knew what to do before they launched new services or marketing campaigns to avoid resource constraints. In 2007, we were doing this with FastWeb, a pioneer of triple-play services in Italy and the first to introduce IPTV with fiber networks spanning across an entire city. They were delivering linear TV and on-demand content such as video rentals. FastWeb asked us if we could apply our predictive technology to their TV and video offerings to see if they could predict what content would be relevant to the customer; acknowledging that the attention span and screen real estate was limited. That project is what eventually became ContentWise, which was officially launched as a version one product in 2010.
Pipeline: It must have been a challenge to personalize linear TV offerings at the time. How did you address to personalize the customer's viewing experience?
Auteri: In the first version, we used an algorithmic technique called collaborative filtering which looks at the consumption event of the customers behavior. Every time they viewed a channel or a band on a channel or rented a movie - or started or stopped viewing - it was recorded by ContentWise which created a mathematical model that looks at users that have similar behavior and then recommends movies, channels, and programs to the customer. Similar to Amazon, if you asked for a recommendation ContentWise would suggest a recommendation based on what people like the customer also liked.
