Designing and transforming an OTT discovery platform for a leading entertainment startup.

NDA Project

It is a social network platform built on discovery that lets users discover content across platforms, rewards them for their interactions, and celebrates the best of cinema.

Context

Unlike in most businesses, the Covid-19 pandemic changed the media and entertainment industry, 2020 turned out to be the year of OTT entertainment, with movie theatres being shut down for most parts of the year. OTT video streaming platforms including the likes of Netflix, Amazon Prime Video, Disney+ Hotstar, Voot, and SonyLIV, among others, gained immense popularity in India.

Our client is a platform that wants to solve the problem of helping users in finding something to watch across OTT platforms, as well as establish a community within the app, making it simple for users to connect with one another and increase social engagement.

So, in order to understand how the existing application works and what issues the user base is experiencing, our team went through a thorough procedure.

Our Approach

To begin the project, our team chose the double diamond technique as our first step in project planning and structuring so that we could move forward with ease. All of the above methodologies were determined during the process, some ideas were added and some were removed.

Double diamond approach

My Role and Responsibilities

Our team had 6 people in total, and everyone was allocated tasks, the task's progress was then reported to the design lead during scrum calls every day.

My role included :

• Cognitive walkthrough
• Heuristic Evaluation
• Desk market research.
• Preparing user interview questions
• Building user persona
• MOSCOW card sort ( team task )
• Feature prioritization ( team task )
• Wireframing
• Visuals for mobile ( Reviews and recommendation section)

Problem Statement

To start with the research the problem context given to us was as follows:

Problem Statement

These were a few of the problems faced by users while having an initial walk-through of the application.

What user have to say ??

⚠️️️️️ DISCLAIMER: Since the client project is under NDA most of the research and visual part cannot be made public. Mentioned below is a glimpse of the entire story.

📌 Cognitive Walkthrough

A cognitive walkthrough is a technique used to evaluate the learnability of a system. Unlike user testing, it does not involve users and, thus, it can be relatively cheap to implement.

A Group of three people was involved in the process of doing the cognitive walkthrough since each person provides a unique viewpoint to the walkthrough and guaranteed that the evaluation is comprehensive.

We looked for answers to the following questions during the walkthrough for each section and the screens within these sections:

• Is the user easily able to perform the task and achieve their desired goal?
• Is their Conceptual Model correct as it aligns with their end goal?
• Are the action performed by the user easily visible on the screen?
• Are users easily able to recognize the labeling of action and perform is correct?
• How many steps into the process does the user encounter a problem or a lack of information necessary to move on?

Cognitive Walkthrough Insights

📌 Heuristic Evaluation

It is a method for finding the usability problems in a user interface design so that they can be attended to as part of an iterative design process.

To understand and analyze the problems, we conducted a heuristic evaluation for all the existing screens.

A few of the insights gathered from the heuristic evaluation are as follows:

• No consistency with the font size, font weight , and colors , their format changes when users switch from one screen to another screen
• Since there is no animation or graphical representation on the screen, the app becomes less interactive and draws attention to potential issues.
• Too many confusing icons with no labels; the same icons for like and recommend.
• There are a lot of hierarchy and structural issues in recommendation screens, profile screens, For you screens, and feed screens.
• The search results do not differentiate between movies, TV Shows, and Actors. All results are displayed in a single list altogether.

Document of Heuristic evaluation

📌 Desk market research

After going through the problem and analyzing it, we tried to understand the market working of the OTT platform. For this, we searched for this information on online mediums like news articles, went through answers on community platforms, and open source surveys.

During the desk research, I also tried to find answers to the following questions

1. What are the most common issues that users encounter when searching for movies on an OTT platform?
2. How do users decide what to watch?
3. How does the content recommendation system of an OTT platform work?
4. Do the present OTT platforms have any way to build social interaction between users?
5. How do users interact with each other on OTT platforms?
6. What is the Average time duration users spent on OTT Platforms?

Insights gathered from the research :

• As of date, there are 500+ Mn users domestically. and 2.2 Bn users globally
• 4 Hrs and 8 Mins is the average time spent by a user on OTT platforms
• Broadband and smartphones are the primary growth drivers and it is estimated to reach 820 Mn by 2022.
• Amongst the users, 54% are metro users and 46% are from Tier 1 — Tier 2 cities.
• From 2021 to 2022, India witnessed a 20% growth in OTT audience.

Content Recommendation System Working

A recommendation engine, often known as a recommendation system, is basically an information filtering system that presents consumers with the most relevant and helpful recommendations. A recommendation engine’s main goal is to improve the customer experience.

Recommendation engines need lots of data — of the right quantity and quality to recommend and recognize patterns

Movie Metadata

The studios or content producers can provide movie metadata, and it only needs to be consumed once. In the absence of such data, content producers can also get metadata from sources like IMDb or similar rating websites/agencies.

User data points

Apart from movie metadata, we also need to use data that describes a user’s viewing patterns, choices, likes, and dislikes, and average watch time.

Here are a few data points or features about a user that is interesting to recommendation engines.

1. Location
2. Language preferences
3. Watch time or watch duration
4. Up/Down Votes

The “Cold-Start” Problem in Content Recommendation

Now, what happens when a new user signs up for the first time on the platform? The platform doesn’t have any information about the user, preferences, etc., so it is quite difficult to recommend content right off the bat. This is called a cold start problem in the recommendation engine

How and what do you recommend to a user you know nothing about?

Now there are two ways to this

1. We can use the user’s IP to geo-locate their location and serve popular content in that geography.
2. Since the platform collects information about users' gender, age, and language preferences at the sign-up stage, we can use that information to make a general recommendation and learn as we go along the user preferences as they engage with the platform.
3. We also add a small form section at the onboarding stage of the application to understand users' preferences in languages, genera, and favorite actors or actresses.

Monitoring the Recommendations

It’s important to collect information regarding the quality of the recommendations while creating a recommendation system. For instance, if we recommend three movies to a user, will the user pick one of them? If so, does the user watch the film for more than X minutes or does he or she leave after a few?

For this Click-Through-Rate (CTR) is a powerful indicator since this helps us show good the recommendations are and what are users spending their time on and this feedback is sent to the AI/ML system to enhance their machine learning database.

Validation: Consumer survey

During the market research, we simultaneously rolled out survey forms to get a better understanding of a larger audience's perspective, expectations, pain points, and needs.

For the existing app, we did a consumer survey which is as follows:

200+ users participated in the consumer survey

Insights for the survey

• 68% of respondents said that they faced difficulty in searching for a content
• 41.9% Relied on Recommendations from Family and Friends
• The majority of people were regular viewers on at least two OTT platforms.
• Binge-watching on OTT platforms is one of the most popular past times.
• 37.1% Relied on independent reviews of the content

Key problems identified

• Unreliable Recommendation –Recommendation from Unverified
  sources
• Scattered content across platforms
• It takes a lot of time to find the proper movie or TV show, which frequently results in time waste and boredom.
• Difficulty in using the app interface often gets confusing and difficult to use.

consumer survey form

📌 Primary research

User interviews were conducted throughout the primary research phase to have a better understanding of the users’ requirements and desires. My task here was to compile a short list of questions to ask the users during the interview.

A few of the questions for the interview were as follows:

1. How often did you watch movies pre covid time?
2. How did you spend your time during the covid phase? Can you brief your daily routine?
3. How many OTT platforms have you subscribed to?
4. How often do you watch movies on OTT platforms and which is your favourite platform?

Along with this, 12 more questions were asked to users in order to have a deeper grasp of their perspectives.

📌 User Persona Building

User personas are idealized individuals with the goals and characteristics of a wider set of users.

Why user persona? 🤔

A thorough understanding of a target audience is necessary for creating good products. A developer can use user personas to answer one of their most important questions: “Who are we designing for?” Understanding the goals, concerns, and motivations of target users allows you to build a product that will meet their needs and so be successful.

📌 MOSCOW

The Moscow is a technique for helping to understand priorities and ensure a common understanding across the different team members. It can help determine a clear set of requirements, and a clear priority for those requirements, by categorizing each user's needs.

• M is for Must have
• S is for Should have
• C is for could have
• W is for would have

Moscow technique was a great help for us in our next step to feature prioritization.

MOSCOW sheet

📌 Feature Priotrisation

The practice of ranking and organizing features in a product based on consumer value, business goals, time and cost, and technological viability is known as feature prioritization.

The feature prioritization is based on two key criteria: the impact the feature will have on the end-user and the effort necessary to execute that feature.

📌 Solution: Wireframing and Visuals

After a detailed review of the problem statement and understanding of users’ perspectives, pain areas, needs, wants motivation, and goals. We created a solution.

We call it 3Ds

1. Discovery: The app would allow users to become acquainted with their own content discovery and monitor how their experience has evolved within the platform.
2. Discussion: Users will also have access to featured reviews and reels focusing on trending content. They can interact with other users whose content and behavior is highlighted.
3. Deals: The users will be able to get the most out of the platform and get rewarded for their experience through various deals, offers, and coupons.

The ideas of wireframing began with pen and paper, followed by low-fidelity wireframes.

Followed by the high — fidelity mockups ( visual designing )

Glims of Figma file to the visuals

📌 Outcomes

After following the design approach, these were the biggest outcomes of the project

1. Ease to Use: The redesign of the application allowed a very accessible and user-friendly design, which in return maximized usability.
2. Increased user engagement: With the intuitive design and enhanced aesthetics, users are spending 32% more time on the app.
3. Increased Retention: With the intuitive design and data flow, users find it easy to consume information increasing the retention rate by 30% on the application.

📚 Learning

This project taught me a lot of things

1. Ask the correct questions and as many as you need to in order to fully grasp the problem at hand.
2. Even though stakeholder meetings were difficult for me, they truly taught me how to convey my views and how to be more understanding of other people’s perspectives.
3. I identified my strengths and weaknesses, realized which part of the process I love the most, and also got a clear knowledge of all the practices.
4. This project taught me the importance of iteration and that every idea should be put into practice using valid reasoning before a final decision is made.
5. While working on the project, I had the opportunity to learn that brainstorming may be enjoyable if done in a group.

📍 This work is under an NDA, so I was able to present only limited information but if you have any questions regarding the project. please reach out to me personally.

You can find me on Twitter and LinkedIn

Or can mail me at ananya.vashist21@gmail.com

Thank you for coming this far.😊