Netflix Case Study: Data Management & Analytics

Question

Task: Context: This assessment gives you the opportunity to demonstrate your understanding of concepts including Configuration Management, Business Intelligence, Big Data, Data Warehousing, Cluster Computing, Data Mining, Machine Learning, Use of Data for AI. In doing so, you are required to select an organisation and then analyse and evaluate how the above-mentioned concepts can be used to solve a real-life problem. Your task is to write a detailed Netflix case study.

Answer

1.1 Introduction to Netflix Case Study
Organization: Netflix
Netflix is a major media streaming platform for home television audiences. It is a video-rental company that gives media streaming along with original programming. Through the Netflix website, the title for television and movie are chosen by the subscribers. It offers media streaming through a subscription model and is currently operating in nearly every country globally. Around $2.5 Million was invested to get the business started in a bigger way(Toshniwal, 2021). The company was first started on 14th April 1998. On the first day, it had 30 employees. It’s almost 23 years when it has started the service as a DVD rental service that would mail DVDs straight to your home. They offered a seven-day DVD and renting it for $4 with $2 shipping charges. To attract more customers, a new titles sale at 30% discount.

In 2007, Netflix made a transition and debuted its streaming service, enabling viewers to watch on-demand TV shows and movies for ad-free. Since then, Netflix has become the world's leading streaming service. Netflix is the simplest and easiest for binge-watching, in spite of being the largest library and the largest online platform in terms of streaming content. Once you sign up with a subscription, a friendly interface pops up that enlists the popular shows to watch. As you start watching the movies it remembers your choices and taste then it also recommends the latest shows to watch.

Reason to choose Netflix:
Netflix has got the major collection of users' data and it can create a detailed profile with meaningful information for which it requires data analytics and that’s the only reason for the success of Netflix. Through certain tools of Big Data and analytics, the original content could be easily greenlighted. Netflix had evolved with the changes in the methods of content consumption whose major impact is on the home entertainment ecosystem that has a powerful player involved with profitability.

What does the business do?
With powerful analytics,around terabytes of data are process by Netflixthat could give information that are meaningful. They use Big data and analytics judiciously as Netflix itself is an analytics company. It provides an incredible insight that allows users to make smart decisions through which they get maximum ROI of whatever choices being made by the users(Team, 2021).

1.2 Concept of Data Analysis
People share hundreds of terabytes of data every day just on Facebook, and over 300 hours of videos are shared on YouTube every minute!

By this, a clear analysis can be made that users are continually giving their data to companies.
Big data analytics means analyzing a large amount of data. Data can help them in faster decision-making and solving problems more efficiently. However, most importantly, data can help them identify new business possibilities that can create more sales and design a better customer experience(Costa, 2021).

Big Data is the success mantra of Netflix.
Big Data is a crucial factor in most online services as it helps attract more audiences and maintains the interest of existing ones by keeping them happier.With Big Data, you get a clearer and realistic picture of your target consumer's preferences and choices.This Data Science helps form the products and services that look interesting and unique and draws the customer's attention.

According to the Director of Global Communications, Netflix has many different versions of around 33 million. By the year July 2018, the world streaming subscribers of Netflix reached 130 million. With this large user base, an incredible amount of data could be gathered through Netflix to make the users happy with their services and decisions.

Netflix tracks the data

• When user pause, rewind, or fast forward
• Content on what day the user watch and mostly it was found users watch shows of TVduring week and in weekends they watch movies.
• The users watch
• Of the time the content watched by the user
• Of the zip code where they watch and
• Of the device, they used to watch
• When the user pause and leave the content and when come back again
• For each rating given per day
• Searches made per day
• Browse and scroll

Through analytics, Netflix comes to know to what extent the content needs to watch by the user so that the user may likely cancel it(Swarge, 2019). Even if each user watches any content for at least 15 hours a month, then there is a 75% less likely chance to get it to cancel. In any case, if the user thinks to drop and bring to 5 hours, then the chances is high to get it to cancel by 95%. Now, as Netflix already has such data there comes a challenge that, what can be done to make users watch every content for at least 15 hours in a month.

1.3 Data Sources
References
Costa, C. (2021). How Data Science is Boosting Netflix. Retrieved 1 April 2021, from https://towardsdatascience.com/how-data-science-is-boosting-netflix-785a1cba7e45

Dixon, M. (2019). How Netflix used big data and analytics to generate billions - Selerity. Retrieved 1 April 2021, from https://seleritysas.com/blog/2019/04/05/how-netflix-used-big-data-and-analytics-to-generate-billions/ James, K. (2020). The Rise of Netflix: A Data Analysis. Retrieved 1 April 2021, from https://medium.com/swlh/the-rise-of-netflix-a-data-analysis-9cbd3e00d736#:~:text=This%20data%20set%20consists%20of,on%20Netflix%20as%20of% 202019.&text=The%20streaming%20service's%20number%20of,from%20the%20same%20data%20set

james, k. (2021). kelroydbjames/netflix-worldwide-analysis - Jovian. Retrieved 1 April 2021, from https://jovian.ai/kelroydbjames/netflix-worldwide-analysis

Netflix Data Analysis. (2021). Retrieved 1 April 2021, from https://www.kaggle.com/chirag9073/netflix-data-analysis

Netflix Research. (2021). Retrieved 1 April 2021, from https://research.netflix.com/research-area/analytics

Patel, N. (2021). How Netflix Uses Analytics To Select Movies, Create Content, & Make Multimillion Dollar Decisions. Netflix case studyRetrieved 1

April 2021, from https://neilpatel.com/blog/how-netflix-uses-analytics/
Smith, Q. (2021). 6 of my favorite case studies in Data Science. Retrieved 1 April 2021, from https://bigdata-madesimple.com/6-of-my-favorite-case-studies-in-data-science/ Swarge, G. (2019). Data Analysis &Visualisation of Netflix Viewing History. Retrieved 1 April 2021, from https://medium.com/analytics-vidhya/data-analysis-visualisation-of-netflix-viewing-history-565cefe288fc

Team, E. (2021).How Netflix Uses Big Data to Drive Success - insideBIGDATA. Retrieved 1 April 2021, from https://insidebigdata.com/2018/01/20/netflix-uses-big-data-drive-success/ Toshniwal, S. (2021).Use of Analytics by Netflix - Case Study. Retrieved 1 April 2021, from https://www.slideshare.net/sakettoshniwal/use-of-analytics-by-netflix-case-study

1.4 Analysis
1.4.1 Original Problem faced by Netflix
In 2006, Netflix suffered a large loss, due to which it made a transition into an online streaming service. While entering the streaming market, it began with a competition for movie ratings(Smith, 2021). However, in 2007 Netflix has changed its business model from a regular DVD rental shop to online video streaming with a great boost of internet users.Competitors of Netflix like Disney and Amazon Prime are far behind it in terms of ratings and users. From a very early stage, Netflix had invested heavily in Data Science.

Although Netflix has protected its user's privacy and anonymized its dataset, yet many privacy issues emerged around Netflix's data. Many researchers figured out the users in the anonymous Netflix dataset by meeting their ratings on the Internet Movie Database. Later, a prosecution was filed against Netflix by four people for the violation of the Video Privacy Protection Act and breach of the United States' fair trade laws.

1.4.2 Implementation of Big-Data
For much long time, Netflix continued its streaming business where it has high viewers' data like gender, age, location, media taste, and more(Patel, 2021). With every information of the customer being gathered, Netflix could dive into the viewers’ minds to get the idea of the user mostly like to watch next before they think of finishing the show or movie.

With time, Netflix could deploy various mechanism and algorithms where the data is used to generate critical insights through some tools and features, namely:

1. Real-Time Recommendation Engine
2. Artwork & Imagery Selection
3. Production Planning
4. Metaflow
5. Polynote
6. Metacat
7. Druid
8. Use of Python

1.4.3 Problem faced during implementation of Big-Data
The Netflix team faced some technical challenges while creating the system were("Netflix Research", 2021):

? To predict a particular outcome various models were grouped together.
? Achieving Weekday Effect and Global Time Effects.
? Distinguishing whether the short-term effects were due to various people sharing the same Netflix account or the differences in a person's moods and taste of shows and movies.

? Unavailability of a video from the point of view of a recommender scheme.
? Identifying, listing, and supplanting the unavailable references.
? It gets very difficult for Netflix team to understand and identify what the user is searching for because the spellings do not match or the size of the search show or movie is usually too short.

? Improving and upgrading customer experience by providing various indexing plans and metrics.

? Optimizing customer experience by allowing different indexing schemes and metrics.

1.4.4 Benefits of Big-Data implementation
With Big Data implementation, the advantage that Netflix has to allow it to know its customers well through the internet. Netflix could drive around 100 million subscribers along with their data to improve the user experience(James, 2021). With the implementation of Big Data, Netflix has now rule the world of stream. It is helpful to decide through Big Data for which program might interest the user and 80 % of the content that the user watch is influenced by the recommendation system. From a customer retention point of view, the algorithms could help Netflix to save around $1 billion a year(Dixon, 2019). A typical member of Netflix loses interest after 60 to 90 seconds before having a choice to watchsomething. It has a review of around 10 to 20 titles. On every homepage, there are around 40 rows each based on the device capabilities and each row has up to 75 videos.

Before 2006 Netflix was a business primarily for a DVD-mailing, later it Netflix Prize got launchedwhere they offered many group with $1 million that can predict the algorithm with best possible way where customers most probably like the movies depending on what ratings that were previously done. In the year 2009, they announced the winning entry. Though they revised the algorithm consistently, the key elements were the same. 80% of the content was streamed on Netflix when influenced by the recommended system.

Through Big Data Netflix could gather a massive amount of data and use them to create user's better experiences.
With the use of Big Data, Netflix could

1. Find the next series of Smash-Hit. It has spent around $100 million on House of Cards has 26 episodes, through which they could market the audience successfully.

2. Personalized the ranker for the video: Each member who has a profile could get the entire collection of Netflix orders in a personalized way. Each member row having the same genre could have a different selection of videos.

3. Top the video ranker: Picks the top personalized recommendations throughout the catalog that focus on the titles having at the top rank.

4. Trending Now: Based on the beacons and viewing history, there is a mix of personalization of the videos trending which capture all events and activities of users on Netflix.

5. Continue Watching: Users could easily sort recently viewed titles and gives estimates whether members could watch it continuously or re-watch in later again or may happen that the user might stop watching some content as there was no more interest than envisaged.

6. Similarity Algorithm of video: Users when watched one video, may land up watching a similar video again. In this case, the similarity ranking may sometimes not be personalized. Where a good estimate would be what a member likes based on what they watched previously.

1.5 Conclusion, Finding, and Recommendation
Things learned from this analysis
The ability to collect data and using it into meaningful information is the only reason for the success of Netflix(James, 2020). With customer retention, Netflix could earn in billions because over 80% of the content is streamed to this platform that has a recommendation system accounts. Netflix has used Big Data analytics tools to greenlight the original content which is derived from the user-based and get viewership and engagement("Netflix Data Analysis", 2021). With the use of Big Data analytics, Netflix could even conduct custom marketing. In this way, Netflix don’t require much time nor much resources to spend so as to show up on promoting. As it is already known to them that people are already interested in it.

Moreover, Netflix also collects feedback from the subscriber through a feedback system by thumbs up/down where the rating system gets substituted. The user's homepage gets customized further with the audience engagement.

Core findings and Recommendations
All data of Netflix could be collected from 151 million subscribers. The behavior of the customer and the buying patterns could be discovered through the implementation of data analytics models. Based on the subscriber has preferred to watch, the information is used accordingly and then the TV shows and the movies are recommended to them.

With the personalized recommendations of Netflix, almost 75% of the viewers' activity is based on it. Many of the subscribers’ profiles are created in detail when the data are collected by Netflix. The data that Netflix collects have more details than those were conventional marketing is created.

Most of the collection of Netflix arethose data that are based on the interaction of customerand to the TV showsresponses. When any user watches a show, Netflix gets to know the timing and date at which the user thought to watch it, and the device that the user uses to watch the show. Whichever show the user pause or watch or even resume are all collected. Whether the user has finish the entire TV show or not, or the time that is taken to finish the show, and many more.

Many a time user might watch a show repeatedly or may give a rating to the content or what search where made and the number of searches made. In that case, Netflix even takes screenshots of those scenes. Through this, Netflix gets the detailed profile of the users and then creates accordingly. With the help of data analytics, Netflix could harness all the user data that are collected into meaningful information. Netflix uses the recommendation algorithm and gives suggestions to the users based upon their preferences, and then the TV shows and movies are suggested to the uses to watch them