EnterTRAINment is a product designed to correct the skewed algorithms of streaming services and improve the accuracy of content suggestions, enhancing a user’s overall entertainment experience.
How might we create an efficient method to correct the skewed algorithms of streaming services and improve the accuracy of content suggestions, enhancing a user’s overall entertainment experience which may suffer when the content suggestions are less accurate.
According to a Cloudwards article, 85% of U.S. households have at least one video streaming subscription, and 60% have at least one paid music streaming subscription, with the streaming industry expected to be worth $330 billion by 2030. The goal of this industry is to attract and keep a large audience by offering a user-friendly service that delivers high quality content and a personalized consumer experience. They accomplish this by providing on demand access to a variety of digital media content, by improving their technology and UI, and by developing algorithms that recommend personalized content to users based on their viewing history, preferences, and other data.
Currently, to make changes to personalized recommendations one must retrain the algorithms by engaging with preferred content one by one or must reset and remake their preferred content algorithms from scratch. They also must do this for each streaming service separately.
However, the problem with this is that it takes a very long time to adjust algorithms using those methods, so many users won’t even bother correcting the algorithms when suggested content does not match their preferences. This leads to lower user satisfaction which impacts engagement and retention rates.
I conducted a survey and distributed it to 22 participants to better understand the streaming service user experience and to uncover suggested content pain points. In addition to user surveys, I also conducted two user interviews to expand on the unique experiences of users of streaming services. Also, to find what role suggested content plays in that experience and what improvements could be made to create a better experience for those users.
- Through survey research from 22 respondents, I found that 96% of streaming service users reported that content suggestions play some role in their overall user satisfaction, yet 77% admit that they made few or no attempts to correct the content suggestions provided.
- Among the biggest pain points for streaming service users is the accuracy of the suggested content. 42% of users said they feel that their streaming services suggested content is only somewhat accurate, or not accurate at all and accuracy of suggestions was mentioned in half of the responses attributed to frustrations with the suggested content
- Majority of participants reported that at least one other user used their account, with only ⅓ of responders saying that they were the only users of their account.
These findings confirmed that a major pain point in the streaming service user experience was inaccurate suggested content by providers. There were several contributing factors to the lack of accuracy but the main one was sharing a profile with one or more users. Through interviews, I learned that while there was a problem with the accuracy of the suggested content impacting the user experience of streamers, most users chose not to address the problem due to a lack of easily available solutions.
Landing on the Solution
A solution to this problem is to create a tool that allows a user to easily create a profile that saves their preferred content choices and then applies that profile to their streaming services with the click of a button.
Since creating the original profile will take time, it needs to be made as simple as possible so it can be applied universally to streaming services independent of what content they are offering at that moment.
If the suggested content algorithm skews away from the user’s preferences for any reason, this would provide a quick and easy method to correct it. There should be a method to select from multiple users in case more than one user shares an account and there should be a way to save iterations of the profile so a user can always go back and apply an older version if they want.
Explanation of Solution
As I worked towards developing my MVP I realized that there were some aspects of the solution that would be better left as features for future iterations of the product. The product needed to be a simple solution that would allow a user to easily navigate through choices so they could select their preferred content and apply it to their streaming service.
In order to simplify the product I needed to remove features such as the saving of iterations of profiles as well as limiting the number of streaming services that would apply. Instead of a user being able to choose any service, they would be given a choice of the top most popular streaming service options that they could apply their preferences to.
The solution is a web based application where someone will create a user profile that has access to their streaming service providers and allows them to navigate through a selection of media where they can choose to interact with the media to better define their preferred content. After defining their profile with accurate content preferences, they will be able to choose which media types they want to apply their user profile to and which users they want to apply it to. These saved user preference profiles will then be applied to the streaming services that the users have access to.
If for any reason the suggested content options from their streaming service becomes skewed over time and no longer matches their taste, they can just log into the website and apply their stored profile to that service again to realign it. They can always go back and make adjustments to their choices or further develop their profiles by defining more media options so suggested content can improve on their streaming services.
Because there will be an initial time investment from the customer to create and define their profile, I wanted to offer several methods for them to navigate the media selection in order to make it a more inviting experience for them. This includes the option to view choices in a scroll view or list view, as well as a search option. A user can also choose to browse alphabetically, by genre, or by star performers of the media type selected.
While I will not be moving forward with the development of my MVP, I have gained a lot of experience that I intend to apply towards my professional development. I am very proud of the solution that I was able to find after spending time learning about the problem space and developing a well thought out product spec.
Product Manager Learnings:
In CoLab I learned the importance of taking advantage of different iterations to improve on designs after receiving feedback..
- adjusting a research survey until the right questions were getting asked
- deciding on features to remove from my MVP to limit complexity
It is important to find the correct research method for your purpose and situation.
- In my research, surveys acted as a useful tool in understanding the habits of streamers.
- Qualitative feedback from user interviews delivered new insights that the surveys alone did not. The ability of the user to expand on ideas and add details to their story provided me with new data that I could use to improve the product.
Building in public is a critical component of product management that I was not previously familiar with. I have practiced building in public for years while boring my friends with product ideas but being able to apply that to social media was very useful. I was able to attract a group of people who are genuinely interested in my ideas and can supply useful feedback.
One of the most important things that I learned was that I had been spending far too much time on the solution space and not nearly enough time on the problem space prior to this point.By focusing on the problem more, I was able to understand the underlying problem that my product is trying to solve.