top of page

Enabling content discovery through AI-Driven content curation

FOR YOU

COMPANY

ZEE5

DEVICES

RESPONSIVE ACROSS MOBILE, PWA & TV

ROLE

PRODUCT DESIGNER

4 MONTHS

TIMELINE

OVERVIEW

Zee5 is a leading Indian OTT  platform known for its vast library of regional content.  "For you" in Zee5 is transforming how you discover your next binge-worthy watch! We've introduced a groundbreaking AI-powered recommendation system that leverages social media trends and Gen-AI technology to create a truly personalized experience.


I focused on the user experience of "For You," designing intuitive interfaces for AI-powered recommendations. This included crafting engaging layouts to showcase social media buzz and award wins, while ensuring clear presentation of titles and summaries.

2.png

CONTEXT

Old "For You" page? More like "For Everyone" page

​​

For You page was like a crowded party: lots of options, but nothing you truly connected with.

​

Zee5's "For You" page was a one-size-fits-all . It lacked personalization, resembled the generic homepage, and failed to engage users. This resulted in low click-through rates and missed opportunities to create loyal viewers. We needed a revamp - a personalized experience with viral content, social buzz, and features like mood-based recommendations to keep users hooked and coming back for more.

UM THE PROBLEM EXACTLY?

No personalization

There was no real estate on Zee5 app which could be perceived as "personalized"

No experience differentiator

“For You” page looked just like home page. The experience was not unique.

a (1).png

No virality factor

Most content becomes boring in 3 days, there was a need for a dedicated place for viral content

a (3).png

PROCESS

image.png

Benchmarking

Detailed competitive benchmarking of similar products that offer awesome personalization.

image.png

Social media sentiment

How to manifest the social buzz that we see in other products in Zee5?

image.png

Objectives

Defined final use-case objectives to see what features do we want to implement.

image.png

I/O Blueprint

To align all the teams we employed systems thinking approach.

image.png

Brainstorming

A team brainstorming workshop with all the stakeholders 

image.png

Decision factors

Based on our research identifying what are key decision making factors and attributes.

image.png

Archetypes

Understanding the user archetypes to understand user needs, drives & blockers. 

image.png

Conceptualization

Final breakdown of components and rail collections to be incorporated.

COMPETITIVE BENCHMARKING

While personalization was our north star, we knew a deep dive into the competition was essential. We analyzed both OTT and non-OTT apps!

By looking beyond the immediate competitors, we aimed to create a truly unique and engaging "For You" experience for Zee5 users.

OTT DOMAIN

To understand successful personalization features in the streaming space - recommendation algorithms, social proof elements, etc.

227_Netflix_logo-512.webp

N​ETFLIX

  • Profiles feature across devices/platforms

  • My Netflix - Notifications (Coming soon, New arrivals, Suggestions); Downloads; Shows you’ve liked; My list; Trailers you’ve watched; Go behind the scenes (Sneak peaks, insider news)

youtube-logo-youtube-logo-transparent-youtube-icon-transparent-free-free-png.webp

​YOUTUBE​

  • Home feed customization

  • Notifications - Choice btw which channels notifications user wants to receive

  • Creator interaction through Community feature

  • Youtube Music - Personalised playlists; Offline Mixtapes

  • Youtube Shorts: Personalised discovery experience; Hashtags

MUBI-logo.png

MUBI

Human curated movie selections over algorithm generated recommendations.

Roku-Symbol.png

ROKU

Roku recommends (A short weekly show hosted by Roku that talks about the weeks best 5 content on Roku)

NON-OTT DOMAIN

To gain insights from unexpected sources! Inspiration could come from social media apps' personalized feeds or e-commerce platforms' curated product suggestions.

imdb_logo.png

I​MDb

  • Recommendations based on viewers' ratings and viewing history.

  • Watchlist feature and User generated playlists by movie enthusiasts

  • Discussion forums

Spotify

SPOTIFY

  • Discover Weekly;

  • The Only You Campaign;

  • Follow Profiles, playlists, etc;

  • Share on your favourite Social Media website

Rotten_Tomatoes.svg.png

ROTTEN TOMATOES

  • Audience and critic ratings

  • Personalised “You Might Also like” recommendations

Letterboxd_2023_logo.png

LETTERBOXD

  • Personalised Movie Diary - Track and rate watched movies.

  • Follow other users' profiles.

  • “Popular” (Within the community)

SOCIAL MEDIA SENTIMENT

Personalization goes beyond user data - it's about understanding the cultural zeitgeist. That's why we analyzed social media buzz around related apps! This helped us see:

​

  • Trending Content: What shows are sparking passionate discussions online?

  • User Preferences: How do viewers describe their favorite content? Are they using humor, excitement, or nostalgia?

  • Regional Nuances: How do social media conversations differ across languages and demographics?

​​

By analyzing social sentiment, we aimed to create a "For You" page that reflects not just individual preferences, but also the broader cultural conversations surrounding the content users love.

Instagram.png
Instagram.png
Instagram.png
Instagram.png

OBJECTIVES

Defining use cases was crucial for our "For You" page refresh. Without clear objectives, personalization efforts could miss the mark. By outlining these objectives, we ensured the "For You" page wouldn't just show content, it would spark conversations and keep users engaged.

01

Showcasing the positive social media sentiment related to a title to increase social proofing

02

Creation of alternate endings for popular TV shows and movies

03

Leverage conversational search on the 'For You' page in the form of an avatar for quick and guided content discovery

04

Personalized recommendations powered by AI/ML models

05

AI-powered Immersive Discovery experience using Reels

I/O BLUEPRINT

We knew a successful "For You" page refresh hinged on clear communication between design and engineering. That's why defining input and output for each personalized feature was critical. Think of it as a recipe! We needed to specify the ingredients (input) - social media data, user preferences - and the final dish (output) - engaging summaries, alternate endings. This ensured features were designed with a purpose, meeting user needs and delivering the desired personalized experience.

SOCIAL MEDIA

ALTERNATE ENDING

RECOMMENDATION

INPUT

OUTPUT

Social sentiment about content on various social media platforms

Content level one-liner to be displayed in recommendation rail

INPUT

OUTPUT

Subtitles and Synopsis

Episode level Alternate story continuation options to be displayed on the page as a widget

INPUT

OUTPUT

 User, content and interaction data

User personalized Recommendation rails

BRAINSTORMING

Head of Design

For you currently stands at 2% of all Home feed clicks → we need to take it to 15%

Product Manager

The Head of the Family ( Not the breadwinner ) controls the Connected TV

Researcher

Let's experiment with things like Vibe Check- ask the user for their mood and recommend content accordingly.

Me hehe :3

1 million DAU consuming more than 10 min per day.
Dedicated placeholder across the platform, which hold potential/viral content in a periodic manner.

Sr. Designer

Conflict - Family time but want to consume personal entertainment, usage through mobile

We should also be able to monetise this while we create the experience, real estate to be given to brands

We need an experience differentiator may be like - Meme generator, leaderboard, gamification, rewards ladder

Engineering team lead

Product Manager

DECISON MAKING FACTORS

To personalize the "For You" page effectively, we needed to understand what drives users' choices. We pinpointed key decision factors: genre, mood, social buzz, critical acclaim, regional preferences, and past viewing history. This data helped us curate recommendations that cater to individual tastes and cultural context.

ARCHETYPES IN ACTION

Based on user archetypes created by the research team, we built user personas and user stories to decode their needs, drives, and blocks. This helped us tailor the "For You" page experience to resonate with the diverse Zee5 audience. Here's a glimpse of the activity:

HOMEMAKER

Age: 30- 50 Years old
Profession: Homemaker
Gender: Female
Location:
Tier 2+ Tier 1

DRIVES

  • Wants to watch her reality shows and TV shows that she regularly watches.

BLOCKERS

  • Tech aversiveness is a blocker for a portion of this audience. Is aware of other OTT platforms, casting, smart TVs etc it will not be a blocker.

  • Would want to watch long forms of content but are unable to do so because of chores at home and involvement of their family.

PSYCHOGRAPHIC

  • Loyalty is skewed, while the user wants to stay up to date, if they are busy and not able to watch a few episodes, then they will take a verbal download from friends and then jump to the latest episode.

  • Feels updated when she watches new content.

  • Feels happy to share what she saw with friends.

BEHAVIOUR

  • She doesn’t have the control over the remote when more people are watching.

  • Decision on what to watch is majorly made by the husband or other older members of the family.

  • Can be AVOD or SVOD (decision is made by husband whether to subscribe or not) If they are SVOD, will move to watching shows on mobile devices when TV remote is overtaken by other family members.

THE FAMILY MAN

Age: 30- 45 Years old
Profession: Service Professional/ Businessman
Gender: Male
Location:
Tier 2+ Tier 1

DRIVES

  • Feels like this is a good way to get the family together when they sit and watch something together.

  • Indian content which teaches value and culture to society

BLOCKERS

  • Explicit content

  • Awkward family moments when watching with family

  • Kids getting more westernized and forgetting Indian culture

PSYCHOGRAPHIC

  • Family oriented, and would prefer content that is not mature on an OTT platform.

  • Goal would be to blow off steam after a tiring day or week at work.

BEHAVIOUR

  • For a service family man will have control over the remote on the weekends as well as post 11PM in the night

  • For a businessman will have control over the remote in the night.

OTHER ARCHETYPES

There are two more Zee5 archetypes for which we did this activity.

THE YOUNGSTER

Age: 18-30 Years old
Profession: Student/ Young professional
Gender: Male/ Female
Location:
Tier 1 + Tier 2

TECH AVERSIVE

Age: 30- 50 Years old
Profession: Blue collar worker
Gender: Male/ Female
Location:
Tier 2 + Tier 3

RAIL COLLECTIONS

The key to a successful "For You" page lies in offering a curated selection, not an overwhelming one. We defined various content rails based on user needs and content attributes. This diverse collection of rails ensures the "For You" page offers something for everyone, leveraging personalization, social proof, and strategic content curation to keep users engaged and coming back for more.

Then after lots & lots of experimentation and multiple rounds of iterations, we created these components for the aforementioned rails.

Because you watched

Set Reminder

Picked from top 10

Games

Sponsored

Multiple Storylines

Hero carousel

Short videos

Continue watching

OUTCOME

Android
As per March 2023, the android built is launched for 20% of Zee5 audience worldwide as part of controlled experiment. Keep an eye out for the update in your Play Store!
Let me know your feedback if you happen to come across this awesome experience soon :)
me after all this
me after all this
me after all this
me after all this

68%

Increase in viewership

IMPACT

This project won the codeathon (in association with Google) at Zee5 and was well appreciated by the entire team and stakeholders.  

 

As per the data we have till now, there has been 68% increase in viewership with a 20% lift in video views per user because of For You. More numbers incoming, will tell you more about it once we connect for a call!

MY LEARNINGS

Data-Driven Design

This project highlighted the importance of translating user data and competitor analysis into actionable design decisions. It helped me improve my ability to analyze user research and competitor features to make  informed interface elements and user flows.

Micro-interactions and Engagement

The explorations on interactive features like mood checks and alternate endings emphasizes the growing importance of micro-interactions in user experience (UX) design. I delved deeper into crafting engaging interactions that enhance user delight and keep them coming back.

Personalization for Diverse Audiences

Zee5 caters to a vast regional audience. This project underscores the need for designing personalized experiences that consider cultural nuances and language preferences. It strengthened my skills in designing interfaces that adapt and cater to a diverse user base.

Stakeholder management

This was a big project spanning across multiple teams including Product, Data, Research, Business and of course, design. It made me more confident in presenting my thoughts and ideas out loud in meetings and giving design a seat at the table.

Thank you, ciao!

FABLAB

Creating a Design Language System (DLS) for FabHotels

bottom of page