Netflix: A case study of transformation for video streaming service
Mobile Applications

Netflix: A case of transformation for
video streaming service

By Jitin Narang / 390 Views / August 22, 2019

Netflix: A case of transformation for video streaming service

With more than 151 million paid subscribers, Netflix is leading the market of streaming services. This American media-services provider was founded in 1997 as an online movie rental service. The company was competing mainly with established video rental stores. Unlike the conventional renting system, Netflix let the users keep the DVDs for as long as they wanted, without any late fee. The renters received the DVDs through mail service. Slowly, Netflix introduced several technologies on its online platform like recommendation system based on ratings. The service became so popular that it reached one million subscribers in the US by 2003.

It was in 2007 that Netflix enabled online viewers to watch the television series and movies online through this platform. By 2013, Netflix had started producing its original content, some of which have become globally popular. Some famous names include House of Cards, Arrested Development, and Orange Is the New Black. In the subsequent sections, we have focused on different technologies and features revolving around the Netflix app.

Netflix Paid Subscribers

Technologies Used

1. Python

The team uses Python through the content lifecycle. Several software systems that are designed to operate CDN (content delivery network) are written in Python. The Open Connect CDN improves the streaming experience of viewers and keeps the cost of Netflix in check. The core team uses Python to automate tasks and explore and clean data. This is also utilized as a source for visualization work.

For tasks like auto-remediation, vulnerability identification, security automation, and risk classification, the information security team depends on Python. To train a few Machine Learning models for personalizing viewer experience, Netflix uses Python application development. This is done within the personalization Machine Learning Infrastructure.

2. Hybrid App

Netflix is a hybrid app, which gives it an edge over native apps. A hybrid app comes with features like easy code portability, low maintenance cost, and fast development speed. As the technology related to hybrid apps keeps on evolving, the streaming service can update its functionalities seamlessly. This technology is suitable for apps in which the users to be online to access the content. Other apps using this framework include Twitter, Gmail, and Instagram.

3. Apps for Production Team

To make the production of content more streamlined, Netflix relies on different software technologies. Recently, it developed a mobile device application known as Prodicle. This app was developed using reactive architecture to help the production crew organize the shooting days throughout different stages of shooting. Everyone involved in the production now stays updated on what’s happening.

Netflix - Better User Experience

Use of AI for better user experience

1. Auto-generation of thumbnails

This is achieved through an online learning framework known as contextual bandits. This framework helps in choosing the most appropriate thumbnail (or artwork) for every member depending on the context. The attributes considered for this purpose include the titles that a member has played, the genres of those titles, language preferences, country, the device being used, and even the day and time.

2. Recommendations

Traditionally, the company relied on a collection of a batch of data. The service collects a set of data related to how the members utilize the service. It is followed by running a new Machine Learning algorithm on this set of data. The new algorithm is then tested against the existing production system. An A/B test helps the team identify whether the new algorithm is better. This is done by trying the same on a random subset of members. Group A members receive the existing production experience, whereas the Group B members receive the new algorithm. The comparison of engagement helps in finalizing whether the new algorithm is suitable for the entire member population. To further enhance the recommendation process, it now prefers online machine learning.

3. Streaming Quality

The team at Netflix understands that browsing/viewing experience varies by the device you are using. Cellular networks may be unstable and volatile, especially when compared to broadband networks. Keeping these challenges in mind, the company relies on adaptive streaming algorithms. Based on the device condition and current network, this algorithm decides the suitable video quality be streamed throughout playback.                                             

Number of registered users under one account

One particular feature that differentiates Netflix from other streaming services is that the former carefully watches all the registered devices under an account. The users can check the list of these devices whenever they sign in. Through the membership plan section, it is possible to replace or remove the existing desktop or smartphones. Presently, a user can have up to four unique systems registered to an account for viewing content simultaneously.

Netflix for Multiple Users

UI Interface

User Interface of Netflix is one of its notable features. You will find plenty of categories that recommend titles under different genres and moods. Unlike regular genres like Comedy and Drama, Netflix does something extra by suggesting titles under categories like:

  • Binge-Worthy TV Dramas
  • Acclaimed Writers
  • Casual Viewing
  • Ensemble TV Comedies

A user also finds side-scrolling thumbnails for most of the titles. Switching to another user from an account takes not more than three clicks. To personalize the user profile, the users can choose the name and upload a profile picture of their choice.

Kids Friendly Profile

Netflix has always remained conscious of the access of its content by a younger audience. This is why it introduced ‘Kids’ profile. Only the safe and compelling TV shows for kids are found in this section. A character navigation bar enables the kids to discover movies and shows among the kid-friendly content.

To add a layer of ‘Parental Controls,’ this streaming service enables the parents to control sensitive control through a password. There is an important lesson one can learn from such functionalities: Understand the pain points of the target consumers and upgrade your app/online service accordingly.

Comparison with competitors

Netflix currently competes with two major players: Hulu and Amazon Prime Video.

1. Netflix and Hulu

Both Netflix and Hulu are compatible with most modern smartphones and tablets. Even the interface of both these services is user-friendly. Some features make Netflix a slightly better option in terms of usability. For example, it has a closed-captioning option for its content since 2014. The parental control of this service is also more robust.

The users can enjoy Dolby Atmos soundtracks for some of the content by Netflix, something you can’t expect with Hulu. Talking about the content, with an investment of more than 13 billion dollars per year on content, Netflix is way ahead of its competitors.

2. Netflix and Amazon Prime

Compared to Amazon Prime Video, Netflix has consistent interfaces in most of the devices it is compatible with. A single member can have multiple user profiles on Netflix, something we have already discussed above. Each profile gets its own set of recommendations.

A unique feature associated with Netflix is its Smart Downloads feature that automatically removes the episodes that a user has already watched. Once the Wi-Fi connection is available, it automatically downloads the subsequent episode, something not available with Amazon Prime Video. The reactive/responsive search of Netflix makes it easier for the users to locate the content they are seeking quickly. A lot of users feel that Search option of Amazon Prime is a little bit cumbersome.

A particular functionality where Amazon scores extra is its X-Ray feature. A user can access bios, soundtrack info, trivia, and cast photos, without leaving the playback screen. Netflix is yet to introduce something like this.

What lies ahead?

The team at Netflix is working on making the experience more personalized. It is currently testing the functionality of personalized tags and text descriptions. The technology related to movie recommendations is also being evolved regularly. It is also possible that many mobile devices will let the users view interactive films seamlessly, a feature currently limited to only a few devices.

Jitin Narang : As CMO, Jitin takes care of global partnerships and alliances for mobile solutions. He leads by example and is often quoted by colleagues and friends as the epitome of team spirit. With an ideal blend of entrepreneurial ability, humility, a hunger for knowledge and unending enthusiasm, he believes that nothing is impossible and is credited for the ‘Ahead’ dimension of TechAhead.

Get in touch

Whether you’re just getting started with your app idea or you’re a multinational
trying to evolve your brand for the future, we’re ready

GET STARTED