AlphaStream vs Netflix: Differences in the Personalization Process

As AlphaStream’s Head of Data Analytics, I spend a lot of my time, especially at conferences, preaching the power of personalization. I know how important personalization can be to digital businesses and the power that it can have.

It’s why 89% of digital businesses are investing in personalized content. That includes companies like Coca-Cola, Sephora, Wells Fargo, and Netflix. They recognize that personalization can reduce acquisition costs by as much as 50% and personalized calls to action are 202% more effective than non-personalized ones.

Because they’re well known I often use these types of companies to explain the importance of personalization and how it can help you succeed in business. However, personalization is a huge topic, and it works differently in different sectors.

One question I heard a lot at a recent conference was, “What makes AlphaStream’s personalization different from other sectors?”

Netflix is such a classic example of personalization that so many people have experienced. It provides a well-known frame to explore the differences in personalization that AlphaStream provides.

Below I’ve gone through some of the ways AlphaStream’s personalization is so much more complicated to shape and control than the equivalent personalization from Netflix.

The End Goal 

The goal of Netflix’s personalization is to help users find something they enjoy watching. They have a single end goal in mind and a relatively linear user journey required to reach it. 

The final goal of personalizing financial experiences is to help users be as informed as possible at any given moment. To offer them everything they might need to make decisions. 

This means it isn’t just one piece of content (like a film or TV show) that we must provide. It’s all the information around a given financial instrument and subject matter that they need to make the most informed decision possible. 

The equivalent for Netflix would be if, in addition to a film to watch, they provided the user with reviews of the films in question, the changes in their review scores over time, the amount of people currently watching the film now, content that will help them understand nuances of themes explored in the films, whether there are any news stories that will affect the enjoyment of the film for yourself and other viewers, and many more very specific details. 

Volume of Inputs 

The volume of inputs taken in by AlphaStream’s financial experience cloud and by Netflix are incomparable. Netflix needs to know what you’re watching and whether you liked it. AlphaStream goes so much deeper.  

We pull in financial information, content, news, and user behaviors. All of these flow into our system to be analyzed in real time, building a full picture and understanding of market behaviors. Then they’re rated and weighted to find out what matters for each individual user. 

Even if we just look at behaviors, we detect hundreds of financial behaviors out of the box, with our SDK you can get creative and capture hundreds more. 

Netflix has far fewer actions to consider. They need to track viewed, part viewed, watched trailer, opened content intro, rewatched, etc. Within the full range of actions its possible to take on their platform, it’s impossible to reach the same numbers of behaviors that we have to track for financial services. 

Catalogue Size 

When you browse Netflix it might feel as if they have an endless number of films and TV shows for you to choose from, but AlphaStream has almost six times the number of financial instruments to consider. 28,000 instruments compared with just over 5,000 titles available in the USA. 

What it means for AlphaStream is that there’s so many more individual items to the track performance of and map behaviors too. So not only are we having to track many more types of inputs and gather more data on each of the individual instruments or films, we’ve got far more of them to track, understand the relationships between, and confer people’s interests in them and other entities likes them. 

Rating importance and determining value of content 

This leads to the issues in deciding the value and importance of these inputs. When you rate a film on Netflix, you give it a thumbs up, that means you liked it. It’s a very simple process for their personalization software to understand. 

We had to create a bespoke behavioral capture software because the world of finance is so complex. The actions taken within it, their meaning, and relations to financial instruments are all unique to that market. Each of these behaviors is analyzed to understand the intent and discover what that means for the markets and for each user. 

We then combine our understanding of those actions with third party data to provide a full, rounded view of financial content, financial market movements, and user behaviors.  

Meta-data 

 The meta-data being considered when choosing a financial instrument is evidently more complex and varied than when you’re choosing a film. But even if we were to say they were even in number and ability to decipher, the financial instruments meta-data still require closer and more constant observation. 

For example, the credits of a film are fixed post-production. Actions and producers won’t change mid-screening. It will still be the same people delivering the content as when you chose to watch it and they won’t add in scenes or delete them after it’s been put into the public sphere. In Financial services, that’s not how it works. Tradeable entities are always adding or retiring products, people at those entities get hired and get fired, the inner workings of your investments will change over time. And they release earnings reports, stock holdings will shift, and equity and debt ratios won’t always stay the same, there’s so many moving parts. 

The Netflix user is considering a fixed set of data within the film whilst the AlphaStream user researches instruments that contain fluid data, constantly changing their views on future performance. 

Influences from the Outside World 

 Netflix lives within a closed system, the AlphaStream financial experience cloud lives in an open one, and it is affected by things entirely out of the control of the platform.  

 Netflix can plan every step of the user journey within their system. They are the authors of every message along the way, but we’re not.  

 AlphaStream has to create a journey with a solid structure but then allow our AI to decide what is most important to the end user. It’s the outside world that then authors the messages they need to see, and it’s us who have to understand which ones to show them. 

 At its most simple, famous actors tweeting about a movie won’t majorly effect whether you’ll enjoy watching it and therefore change the experience you need to get from Netflix.  

Meanwhile, famous faces in the business world (like Elon Musk) tweeting about financial instruments and products that you might invest in will materially impact the value of those things.  

Even if Netflix was affected in a more meaningful way by external factors, the risks and rewards involved in financial services make them far more important. Imagine if Netflix was a broker and the film you’re looking at is a financial instrument. Each film would need to be rented individually and that rental price would change from second to second. Popularity would be determined by a huge ecosystem of professional and amateur contributors and the price of the rental at any given time. If you missed some news about the movie you were going to rent, the price could change and cost you hundreds of times the price you’ve paid. 

Fluctuating Valuations vs Fixed 

From the moment of theatrical release, when a film hits the market, it’s market value will have a steady decrease over time. Compare this to a financial instrument’s valuation which will fluctuate and has so many ways to add value to it. These could be technical value creators or fundamental ones.  

The rating or value of a movie might change a little over time as more people see it and add their ratings into the mix, but it’s unlikely to swing hugely one way or the other over time. Yes, there are films like The Room that become classics because of how bad they are, but for the most part a bad film remains a bad film and a good one remains a good one.  

The consensus on a financial instrument is much more volatile. Instruments can fall in or out of favor within a single news cycle, so it’s not enough to slowly allow your reviews to build, you must constantly keep them under review. 

Financial metrics and signals of future changes in value can shift massively in real time. They’re affected by all of the things that we’ve talked about above, from collective and individual behaviors to changes in meta-data to external influences. This means you can’t only consider the impact of long-term performance of an instrument; some users will also need analysis and to be kept informed of signals as they happen real time so they can take advantage. 

Content and Data Generation Over Time 

Almost all films hit their peak popularity within the first 6 months of their existence. Not to say they become irrelevant after that, but the interest in them, conversation around them, and content generated towards them is likely to have settled down to a simmering level.  

Popularity in financial instruments doesn’t see the same decline over time. The average age of an instrument on S&P 500 is 21 years. Over that time they’ve continued to generate high levels of content, data, and information around them, keeping the conversation around them bubbling over. Any of this content could be relevant to and interacted with users at any given time, and therefore has to be considered when you’re providing a personalized experience. It’s yet another way that the potential data inputs into the AlphaStream system are wider than they are for someone like Netflix.   

And Many More 

These are just some of the many ways that AlphaStream and Netflix are different when it comes to the personalization process. I haven’t even gone into subjects like the inherent differences brought by stricter regulations. 

So, whilst I use Netflix as an example of the power of personalization in disrupting a market that doesn’t mean we’re the same. Personalization in the context of financial markets must deal with a much larger set of issues and tensions, and the way that personalization plays out is driving towards a different goal. 

I love talking about the power of personalization and am always happy to discuss how AlphaStream not only deliver better financial experiences but also provide incredibly powerful data and insights to our customers, so get in touch if you’d like to hear more. 

If you’ll be at Benzinga Conference NYC or Finance Magnates in London, make sure to swing by our stall and say hi.

Written by
Ricardo Perez

Might Interest You

Elon Musk’s Tweet Throttle: The Consequences for Retail Traders

Read more

Elevating Financial Experiences: The Unseen Power of a Structured Content Approach

Read more

Privacy and personalization on mobile apps

Read more

Elon Musk’s Tweet Throttle: The Consequences for Retail Traders

Read more

Elevating Financial Experiences: The Unseen Power of a Structured Content Approach

Read more

Privacy and personalization on mobile apps

Read more