"Hands up if you’re a subscriber to Netflix?" I asked at a conference in New York recently, and very nearly every single hand in the room went up. I then asked who was a Netflix stock holder, and a few hands were raised, mine included. So, for full disclosure, I am a stock holder of Netflix. This article isn’t intended to sell or promote Netflix services, but they have created a very successful business and it’s always interesting to look at what asset management can learn from other industries.
First, let’s look at some Netflix history. Founded by Reed Hastings and Marc Randolph in 1997 to offer online movie rentals, apparently after Reed was fined $40 for a late return of a Blockbuster rental of Apollo 13 (although there is debate about the accuracy of this story, but it sounds nice, so we’ll stick with it). In 1999 they moved to a subscription based ‘all you can eat’ model allowing you to rent an unlimited number of DVDs each month. After you had returned a DVD, they would simply send you the next one on your watch list.
In 2000, they introduced Cinematch, their first foray into the world of algorithms and software. Cinematch was the first Netflix ratings system and was designed to recommend new movies based on your viewing history and ratings. It was the first time Netflix put usage data to work to see if it would increase engagement levels (they later, in 2006, created the Netflix Prize of $1m to see if anyone could develop a better recommendations algorithm in one of the first examples of crowdsourcing a tech solution from the public).
Also, in 2000, rather amazingly, Netflix reportedly offered itself for sale to Blockbuster for $50m. Blockbuster chief, John Antioco, turned the offer down in one of the biggest examples of lack of vision in corporate history. Netflix offered to run the online brand of Blockbuster in return for them to run the Netflix brand in stores. They were laughed out of the Blockbuster offices and the rest is history.
In 2002, Netflix listed on Nasdaq, selling 5.5m shares at $15 each, raising $82.5m to invest in technology and growth. At the time it had 600,000 subscribers. They grew their subscriber base rapidly to 2.4m in 2005 and as of January 2018, they have over 114m subscribers.
2007 marked a pivotal year for Netflix. It’s the year they launched online content streaming. This mode of service is always something Netflix believed in, according to founder Reed Hastings. In an interview with Inc.com in 2005, he said: "Movies over the internet are coming, and at some point it will become big business,". "We started investing 1% to 2% of revenue every year in downloading, and I think it's tremendously exciting because it will fundamentally lower our mailing costs. We want to be ready when video-on-demand happens. That's why the company is called Netflix, not DVD-by-Mail."
In 2008 another big decision had a huge impact on Netflix. They decided to migrate their technology to Amazon Web Services. The project was launched after a major outage of the service which was caused by a database failure preventing any DVD shipping for three days. According to Yury Izrailevsky, vice president, cloud and platform engineering, “That is when we realized that we had to move away from vertically scaled single points of failure, like relational databases in our datacenter, towards highly reliable, horizontally scalable, distributed systems in the cloud.”
The AWS migration project was fundamental in the global success of Netflix, it was a huge project and was only completed in 2016. An eight-year journey to the cloud that has set the business up for repeatable, efficient, and global computing power to deliver their service to anyone, in any location.
Finally, in 2013, Netflix released its first original content. House of Cards was released in February 2013 and marked a move away from purchasing libraries of content to stream online, to producing in-house original content to compete with the major television and movie studios.
What did Netflix get right?
Subscription model: The business model is based on monthly subscriptions with recurring revenue. This provides great income visibility and costs can be managed against it. This regular income allows the business to invest heavily in new content. Interestingly, StatPro has always had a recurring revenue business model, even before we migrated to the cloud and SaaS. Our licenses have always been annual recurring agreements, which again, provides revenue visibility which to base investment and costs on.
Streaming content versus delivery: This may have always been the vision, but the decision to launch the streaming service in 2007 and gain first-mover advantage has been crucial to the overall success of Netflix. Having the vision to see that the internet would one day be able to deliver video content, to any home, has resulted in today’s success. This has also focused the company on technology and how to stream content as efficiently as possible. New ways of optimizing video are developed every year allowing the same HD content to use less bandwidth over each connection. This comes in handy when you’re streaming over 125 million hours of video every day.
Original content versus re-selling other peoples: A major decision and one that has lead to such huge numbers in subscriber growth. Netflix knew that money spent on other people’s content was ultimately wasted. The way to hook new subscribers was to have exclusive content available nowhere else. This required serious investment and needed to be a hit early to avoid being labelled as a dud. They were successful with House of Cards and were able to follow up with Orange Is The New Black, Narcos and Stranger Things. These shows built large followings and thanks to social media channels, were able to attract new subscribers in their millions.
Migrating technology to the cloud: This decision has allowed Netflix to go global. It was originally done to resolve single points of failure within critical apps such as billing, but as Netflix continued with their migration to Amazon Web Services, the company realized this was the way to scale globally with an efficient and repeatable service. Like many businesses, Netflix has large monolithic applications with large databases attached. These were scaled vertically with larger and larger servers to handle volume, but eventually went bang when the underlying architecture couldn’t handle any more. By switching to horizontally scalable, highly-reliable and distributed systems, the company has been able to deliver a more reliable service (critical when downtime gets global press coverage) and have the computing power to collect and analyze usage data to help improve the service and guide future content investment decisions.
The ultimate in self-service
So, what has Netflix built? It's built the ultimate in personalized self-service experience for TV entertainment. You can consume content wherever and whenever you want. For most shows, there is no official schedule, you can watch all the content at once. The term ‘binge-watching’ has come about because of Netflix, you can binge-watch entire seasons if you like.
You consume the content you care about; no advertisements and no other shows you don’t care about getting in the way. It’s clean, direct and focused on your choices. You can access content from anywhere, all you need is an internet connected device. You can use multiple devices and you’re not restricted in how you consume the content, it’s your choice. A TV, a laptop, a tablet, or a phone – all are supported.
The interface is built around the content you care about. Recommendations are based on data and you still have the option to browse the entire library if you wish. It’s a very personalized presentation of content. It knows I watched Stranger Things and it knows I like it, so it recommends other content that might appeal to me. It’s the old Cinematch system on steroids.
While presenting a very personal interface with content to match, it’s all controlled. Everything I see is controlled by Netflix and the company is in complete control of what content is made available to me. I’m not accessing pirated material, there are no copyright issues. It’s controlled self-service but made to feel open. This is especially important when we consider what asset managers can learn from the Netflix experience.
How does this translate to asset management?
Asset managers are under pressure to deliver more information to more stakeholders than ever before. Not only is it crucial to deliver returns, but it’s also crucial to deliver the information on how those returns were made, and what risks were taken to achieve them. The job of this data collection, analysis and delivery falls to the middle office within an asset manager. Often seen as a cost centre to operations, with the right technology and vision, there is no reason why the middle office shouldn’t be a real source of high level value to the business. The middle office sees what is happening across the business. If it can’t explain what is happening to senior management, then questions may be raised about trading activities and if they should be allowed to continue. Regulators also want to see what is happening in more granularity than ever before as well. The front office trading desks want to see the analytics of trading decisions to help with decision-making and to stay compliant. Sales and marketing teams need to explain the investment strategy to clients and prospects. They need information in a certain format, so that it can easily be explained.
All these stakeholders want information in a way that makes sense to them. They all need information packaged-up and ready to consume. They also want to make requests for occasional customization. An RFP may come in that requires a specific report or analysis. A client may ask to see something over a custom time period or with non-standard metrics. Senior management may want to drill into the details on a specific strategy or portfolio manager. All these demands add to the pressure felt in the middle office. How can they efficiently produce information for all these stakeholders, every single day, and still keep up with urgent ad hoc requests? The answer is controlled self-service, so stakeholders can consume the information they want, when they want it.
Some asset managers have started this process but uploading a static and generalized PDF to a web portal is not a digital strategy and it is not self-service. Many people in the middle office are reluctant to provide certain tools that they feel will lead to more questions from stakeholders and more work in the long run. This is a fear of a ‘wild west’ out of control self-service, where anyone can login and create analytics they don’t understand and distribute to anyone they like. This is certainly not a desirable outcome but pushing forward with a self-service strategy doesn’t have to mean a loss of control.
A monthly only process is the equivalent of a restricted TV schedule. Why make people wait until the end of the month? It only leads to frustration and risks alternative unofficial sources of information being used.
Restricting the contents of a report by making everything static is the equivalent to a small local Blockbuster video library. Stakeholders might want to see more than is available, but how can they see what else might be of interest to them if everything is locked down in a static PDF? Without the ability to gather feedback on the analysis and reporting being made available, the middle office doesn’t know what is a useful output that should be expanded on and what is a waste of time.
Not getting analytics on the analysis and reporting is the equivalent of blind video lending from Blockbuster. If I walked into a Blockbuster store (hypothetically, there are none left) and hired four video tapes, then returned them in five days, Blockbuster has no idea which movies I watched. It could be one, or it could be all four. They also have no idea which movie I liked and which one I hated. This translates to the output of portfolio analysis and reporting from the middle office to all those stakeholders. Getting this feedback loop is a huge advantage of a self-service strategy. You can see which reports and analyses are being looked at, you can gage feedback and then expand on what is popular.
Sticking with legacy on-premise applications and technology restricts the ability to innovate and creates operational risk. Netflix suffered operational risk with legacy on-premise technology which manifested itself in three days of total system outage. This operational risk exists in almost every asset manager in the world due to the natural growth in systems over the years. Many production systems pre-date Netflix’s decision to start streaming in 2007. It’s no surprise that innovative features are hard to deploy.
What can asset managers do?
Asset managers can help their middle office add real business value by adopting a self-service strategy and creating a place where all stakeholders can help themselves to portfolio analysis content that is designed specifically for them. Each stakeholder requires subtly different views of the information. Working with them to design these views will result in higher engagement levels with fewer questions and ad hoc requests.
- Design views that are tailored for the audience, but don’t restrict what is available
- Suggest other views or data that the audience might find useful
- Gather feedback on analysis to see if it should be expanded on
Centralizing control of content is essential to manage stakeholders and maintain the authority of the data and analysis. User management is important so it’s clear who has access to what. Creating user profile groups makes this easier and users can become members of profiles. Having granular options for what data and analytics can been seen by each user group allows for the creation of tailored views. This level of control also needs to flow through to the interface. Being able to design an analytics dashboard layout and configuration is the final step in building something that is very personal to the end user but controlled behind the scenes by the system administrator.
Utilizing the cloud for scale and efficiency is the best way to deliver such a service. Software as a Service comes with the flexibility needed without the local IT headaches of legacy on-premise systems. The underlying elements of the technology such as hardware and operating systems should be abstracted away from the administrators and users of the system. Upgrades should be seamless and frequent, and technologies should be able to integrate with each other through Web APIs to bring data together.
Netflix saw the benefits of elastic cloud computing and being able to handle spikes in volume and load on the system. This technology is available to asset managers today. Month-end processing shouldn’t take hours-on-end with an elastic system. Calculating fixed income analytics on large portfolios with huge benchmarks shouldn’t bring things to a standstill. Cloud-based systems should automatically scale up as needed then scale back down during normal operations. Distributed systems should be able to handle disruptions and continue to be available during component failures. Netflix deploy ‘chaos monkeys’ (freely available to other developers), bits of code they let loose in their platform to cause failures and issues, so they can see their impact and engineer the platform to handle failure without causing the service to fail. Again, this technology is available to asset managers who are embarking on a cloud migration strategy.
So, don’t be a Blockbuster Video.
Explaining who Blockbuster are to a millennial is hard enough.
“You had to go outside and walk to a store which had a couple of hundred movie titles available to rent. You paid to rent a video tape, but you got fined if you brought it back late, like a library fine.”
“What’s a video tape?”
“Well it’s a cassette, about the size of a book that you put into a machine…”
“What’s a library?”
“Oh, forget it.”
Embark on a self-service strategy with new technology and deliver real value to your business through personalized but controlled self-service portfolio analytics. And don’t forget, the Netflix stock price is up over 8,000% since they launched the streaming service back in 2007.