BUILDING A CURATED STREAMING PLATFORM FOR YOUTUBE SHOWS

Uncovering the Problem

How do you discover high quality episodic shows on YouTube if you don’t know they exist? 

This was the question that I started asking myself while I was working at YouTube. I had been assigned a research project to analyze what types of serialized episodic content (narrative and non-narrative) existed on the platform, in order to determine what types of shows were thriving on YouTube and why. 

On the surface, this seemed like a simple project, but it would end up leading me down a deep unexpected rabbit hole.

I started out by defining a show as a set of multiple videos (episodes) that share the same format and a common overarching title (show name). Narrative shows usually require that episodes be watched in a specific order, while non-narrative or unscripted show episodes can usually be watched in any order. Shows can be seasonal, but not all are.

One of the first things that astounded me was that YouTube had no data on “shows”. YouTube knew everything there was to know about videos, playlists, and channels, but “shows” were not an entity that YouTube kept track of in their database. 

This became a problem when I tried to answer simple questions such as “how many shows are there on YouTube’s top 100 channels?”. Channels did not equate to shows, since channels could contain multiple shows (or none at all). Playlists also did not equate to shows, since videos in a playlist do not always meet the above definition.

In the end, I had to narrow the scope of my research and collect all the data I needed manually. I concluded my research project by highlighting that there was a vast amount of high quality episodic content on the platform that could benefit from additional means of discovery. However in order to do so, it would have to involve the classification of what was a ‘show’ or ‘episode’ by YouTube, product features relating to shows, seasons, and episodes, and creator education about all of the above.

Some of these conclusions led to limited product testing of show-related features, which were later used when YouTube produced their own episodic content under their YouTube Originals program. However, a more thorough push to classify and highlight episodic content never became a priority.

COMING UP WITH A SOLUTION

Long after my research project ended, I couldn’t shake the realization that YouTube’s biggest episodic gems remained hidden. Why wasn’t anyone fixing this? That’s when I decided to take matters into my own hands.

I first tested my shower-thought idea by taking a screenshot of the Netflix interface and overlaying some thumbnails from some popular YouTube shows. After sharing my concept around and getting some “hey that looks pretty neat” comments, I decided to take the idea one step further and create a prototype using Squarespace.

The prototype did a good job of visually demonstrating the concept, but the way it functioned was very, very simple. Each “show” was a blog post that contained an embedded YouTube playlist containing all of the episodes for each show, and the show thumbnails were the featured images for each post. 

I manually curated a hundred or so shows, and categorized each of them with tags such as comedy, education, or food. I created category pages for shows matching each tag, and created a featured show section at the top of each category page. Despite being limited in functionality, I now had a MVP that I could use to gather some feedback. 

In order to reach potential users, I created an ad template and used it to make GIFs that promoted specific shows and general categories on the site. I posted these ads to Instagram, Twitter, and reddit, and started to receive some positive feedback from random internet users.

As I iterated on the MVP, my efforts caught the attention of Next 10 Ventures, a VC focused on digital creators. They saw the potential in the concept and approached me with an opportunity: Would I be interested in scaling this idea within their company?

I ultimately joined the company to lead their Research and Development team, and Next 10 Ventures agreed to fund the development and marketing of Series of the Week.

CREATING THE DATABASE

After hiring several talented engineers and a ui/ux designer, we began the process of rebuilding the platform from scratch. One of our first steps was to establish a proper database, which we did by using Airtable.

Using Airtable allowed us to dynamically display content from Airtable on the front end using their versatile API, and gave site administrators and curators an easy way of managing content updates without having to create our own backend CMS.

We established a system to categorize and feature content for display on the front-end by creating a taxonomy of tags and categories.

Tags were applied to each show when entered the database. The tags were selected based on the show format (e.g. talk show), the genre (e.g. comedy), overarching topics (e.g. food), and other factors that we might want to use for further categorization (e.g. nostalgic).

We created tables with grouped sets of tags that formed categories (ex: education), and sub-categories (ex: history). We linked the contents of these tables to the category pages, giving us dynamically displayed show content by category on the front-end.

Lastly, we created a “featured content” table, where we selected content to be highlighted by giving it a “featured” tag and specifying a date range for it to appear on either the homepage, a category page, or both.

Front End Development

While we were creating the backend, we also began overhauling the front end design. Our goal was to streamline the user experience, introduce new features, and create a unified recognizable brand.

We put a lot of thought into how to make Series of the Week users feel like part of a community. We wanted the experience of using the platform to feel familiar, like Netflix, but imbued with a quirky personality of its own.

We defined a color palette, selected fonts, and created a new logo, and then had a bit of fun propagating design elements here and there throughout the site to reinforce the distinct personality of the platform.

We created pages for each show with featured artwork, navigation menus for seasons, related series, and episode sorting options. To display the episodes for any given show, we used the YouTube API to grab all of the videos within a given playlist url, and displayed them in an intuitive grid format. For the watch experience itself, we added a responsive video player.

We also set up user accounts with customizable profile pages, where users could add pictures and banners, curate their favorite content, and show off achievement badges earned through site behavior.

Curation and Approval Process

During the site development, we created a curation process for identifying new show content based on specific keyword searches and strategies built on my show research, and designed an approvals process for curators to follow when adding new content to the platform. 

We hired a curator who would identify new content through our leads, categorize and tag the content, check the quality of the artwork, and recommend it to be featured. Through this process, we curated 600+ shows into more than 30 categories and subcategories while we completed work on the software development.

(Aborted) Beta Test

The goal of the version of the platform that we were building was to conduct a beta-test in order to gather user feedback and behavioral data from end users, and to have a product that could be demonstrated to industry influencers and investors for further investment.

Unfortunately, right at the time that we were ready to begin testing and industry outreach, funding for the project became no longer available. 

We had to shelve the project.

As an entrepreneur and project leader, I have always taken to heart that you learn more from failure than success, and so despite the project’s abrupt end, I walked away with invaluable experience: leading a cross-functional team, designing scalable content taxonomies, and building a brand from scratch. More than anything, this journey solidified my passion for creating products that solve real discovery problems.