On March 7th 2019, TF1 hosted the highly anticipated Paris Video Tech meetup, and our VP
Engineering Randa Zarkik took the floor to speak about recommender systems, sharing
the stage with Manuel Bergerot, Senior Sales Director EMEA at Nice People at Work.
Nice People at Work and Spideo have been partners for years, combining both companies’
expertise to deliver the best insights on users, their video consumption, and prevent churn.
This time, we unveiled ways to use data and help subscription-based video platforms
convert freemium and trial users into premium subscribers and engaged users.
Randa gave exclusive insights on the different approaches to recommendation
focusing on the the importance of data within content-based approaches. She showed
how Spideo is using data to build trust with our customers’ users, acting as the opposite of
a black box algorithm, and providing them with a fully transparent personalization
In a nutshell: “Explanations might matter more than predictions” by Xavier Amatatrian
Manuel pursued by presenting engagement metrics that they track, allowing to
understand additional churn factors such as quality, relevancy of catalog etc.
In a nutshell: “Keeping a customer happy is cheaper than gaining a new one”
Paris Video Tech organized a great event where professionals of all the online video
community gathered to exchange about the industry! We would also like to thank Nice
People at Work for being a great, supportive and complementary partner. And, last but
not least, an important point; Randa was the first woman speaker at the Paris Video Tech.
When that happens the day before International Women’s Day, or any other day, we have
nothing but positives to say about it.
Are you looking for the next generation of GDPR-friendly personalization solutions for
your video, VOD, OTT, platform?
Contact us now to schedule a meeting and be shown a demo on how Spideo can improve
We would love to hear from you:
Randa Zarkik: firstname.lastname@example.org /
Sarah Rashidian: email@example.com
*Comment utiliser au mieux la data dans la vidéo ?