際際滷

際際滷Share a Scribd company logo
Building a personalized UI for VOD
services
hello@tanktop.tv
http://tanktop.tv
@tanktoptv
息 2014 Tank Top TV Ltd
Mary Meeker highlights Netflixs UI as
a great example of personalization
Building a personalized VOD UI
Any VOD operator can now build a personalized UI
based on Tank Top TV technology.
Heres how.
Well start by looking at an example UI structure
and then look at how our cloud-based service
enables building it.
息 2014 Tank Top TV Ltd
Hero content
Planned viewing
Popular
Recommendations
Recommendations
Basic structure The UI is broken into sections, and
each section is tailored for the
user.
Section ordering is algorithmically
determined for the user based on
behavior.
息 2014 Tank Top TV Ltd
Algorithmically picked from
limited curated selection of
hero content (original
programming, new
blockbusters, special
promotions)
 Fits with users viewing
pattern
 Hasnt seen (recently)
 Random element
Hero Content
Planned viewing
Popular
Recommendations
Recommendations
Hero content
息 2014 Tank Top TV Ltd
 Series the user is part-
way through
 Watchlisted items
 Prioritize items that are
expiring soon or newly
available
 Viewing pattern
considerations
Omit planned viewing if the
user hasnt set any up yet
Hero Content
Planned viewing
Popular
Recommendations
Recommendations
Planned viewing
息 2014 Tank Top TV Ltd
Popular items, filtered for
user preferences &
reordered according to
likelihood that this user is
interested
Weight towards currently /
recently popular
Helps address Fear Of
Missing Out
Hero Content
Planned viewing
Popular
Recommendations
Recommendations
Popular
息 2014 Tank Top TV Ltd
Hero Content
Planned viewing
Popular
Recommendations
Recommendations
Pure algorithms
 Because you watched
 You might like
 Your friends like
Catalogue-driven
 Newly available
 Expiring soon
Curations & lists
Recommendations
息 2014 Tank Top TV Ltd
Hero Content
Planned viewing
Popular
Recommendations
Recommendations
Learning from behavior
Section ordering changes
according to what works 
scores are adjusted according to
 viewing patterns
 business rules
 randomization
息 2014 Tank Top TV Ltd
Personalization platform
User profiles &
events Catalogue
&metadata
Business
rules
services
Personalized
recs & listings
Abstracted personalization services
for your catalogue & user base 息 2014 Tank Top TV Ltd
Personalization platform
Recommender
system
Watchlists &
viewing history
Personalization
UI
HTTPAPIs
Personalized
suggestions
User events
Catalogue
Metadata
Curations
Business rules
User profiles
User history
Time &
availability
suggestions
Contextual
filtering &
ordering
Social connections
Service
Popularity
Beyond VOD UIs
Contact us for our latest thinking on using
personalization for
 customer emails promoting video content
 UIs that combine VOD, live TV and PVR
hello@tanktop.tv
息 2014 Tank Top TV Ltd

More Related Content

How to build a personalized interface for VOD

  • 1. Building a personalized UI for VOD services hello@tanktop.tv http://tanktop.tv @tanktoptv 息 2014 Tank Top TV Ltd
  • 2. Mary Meeker highlights Netflixs UI as a great example of personalization
  • 3. Building a personalized VOD UI Any VOD operator can now build a personalized UI based on Tank Top TV technology. Heres how. Well start by looking at an example UI structure and then look at how our cloud-based service enables building it. 息 2014 Tank Top TV Ltd
  • 4. Hero content Planned viewing Popular Recommendations Recommendations Basic structure The UI is broken into sections, and each section is tailored for the user. Section ordering is algorithmically determined for the user based on behavior. 息 2014 Tank Top TV Ltd
  • 5. Algorithmically picked from limited curated selection of hero content (original programming, new blockbusters, special promotions) Fits with users viewing pattern Hasnt seen (recently) Random element Hero Content Planned viewing Popular Recommendations Recommendations Hero content 息 2014 Tank Top TV Ltd
  • 6. Series the user is part- way through Watchlisted items Prioritize items that are expiring soon or newly available Viewing pattern considerations Omit planned viewing if the user hasnt set any up yet Hero Content Planned viewing Popular Recommendations Recommendations Planned viewing 息 2014 Tank Top TV Ltd
  • 7. Popular items, filtered for user preferences & reordered according to likelihood that this user is interested Weight towards currently / recently popular Helps address Fear Of Missing Out Hero Content Planned viewing Popular Recommendations Recommendations Popular 息 2014 Tank Top TV Ltd
  • 8. Hero Content Planned viewing Popular Recommendations Recommendations Pure algorithms Because you watched You might like Your friends like Catalogue-driven Newly available Expiring soon Curations & lists Recommendations 息 2014 Tank Top TV Ltd
  • 9. Hero Content Planned viewing Popular Recommendations Recommendations Learning from behavior Section ordering changes according to what works scores are adjusted according to viewing patterns business rules randomization 息 2014 Tank Top TV Ltd
  • 10. Personalization platform User profiles & events Catalogue &metadata Business rules services Personalized recs & listings Abstracted personalization services for your catalogue & user base 息 2014 Tank Top TV Ltd
  • 11. Personalization platform Recommender system Watchlists & viewing history Personalization UI HTTPAPIs Personalized suggestions User events Catalogue Metadata Curations Business rules User profiles User history Time & availability suggestions Contextual filtering & ordering Social connections Service Popularity
  • 12. Beyond VOD UIs Contact us for our latest thinking on using personalization for customer emails promoting video content UIs that combine VOD, live TV and PVR hello@tanktop.tv 息 2014 Tank Top TV Ltd