A social video discovery platform focused on human video curation
Discovering a Problem
Where do you go to discover videos that your friends like?
People love sharing videos with each other, but existing online video platforms aren’t designed around human to human recommendations. Platforms like YouTube and TikTok focus on video creation and algorithmic video discovery, but don't have many features aimed at video curators and social video discovery. Reddit and Twitter both allow users to share videos, but are also not focused on video content or curation.
Designing a Solution
As an avid video curator myself I wanted to solve this problem, and so I designed a social video discovery platform that singularly focused on human to human video curation named Malk.tv.
Key Product Features:
Sharing
Curator Profiles
Building a Prototype
I decided to build a prototype to test the viability of my concept, in order to determine:
• If real users found the idea of human-driven video sharing compelling
• If the user behavior of real users matched my theoretical vision
• What kinds of content would be shared by real users
• Any unexpected use cases and user behavior by real users
• Interest in the further development of of my concept
That meant that I needed to build a prototype that could:
• Present an attractive, responsive front end ui
• Support user sign up / login and individual profiles
• Support user created posts and hashtags
• Support user follow, like, and comment behavior
• Support post categorization, filtering, and notifications
• Allow me to track and analyze user behavior
In order to accomplish this, I selected Webflow as my front-end platform, Airtable for the backend database, Memberstack as a platform to handle user accounts, and Make.com (formerly Integromat) to handle data automation.
Tools Used:
The Platform
Easy Video curation
Easily share videos that you enjoy from YouTube or Vimeo: paste the url, add a caption, category, and tags - and that’s it.
Curating Discussion
Set the mood with a video, prime the conversation with a caption, and find out what others have to say about your favorite videos.
Social Video Discovery
Browse videos shared by real people by exploring categories, trending hashtags, and by following unique curators.
Easy Video curation
Easily share videos that you enjoy from YouTube or Vimeo: paste the url, add a caption, category, and tags - and that’s it.
Curating Discussion
Set the mood with a video, prime the conversation with a caption, and find out what others have to say about your favorite videos.
Social Video Discovery
Browse videos shared by real people by exploring categories, trending hashtags, and by following unique curators.
The Alpha Test
Once the prototype was complete, I enlisted the help of friends and family to conduct a limited alpha test to gather data on how real users used the platform.
Over the course of several months, users were encouraged to create profiles, post any type of video content that they wanted to share, engage with posts (like/comment), and to follow other users. Users were also highly encouraged to share their feedback on what they enjoyed and what they felt could be improved.
Key Learnings
Content:
Music was by far the most popular category with 181 posts, with comedy coming in second at 46 posts. Beauty / Fashion and Sports were the least popular. This was well within my expectations, as most people enjoy music and comedy, while other categories reflected more individual interests.
Some users created multiple accounts, each with specific themes. Some focused on specific types of music (ex: emerging artists, lo-fi), others on niche interests (sandwich recipes, vegetarian recipes, plants / gardening).
Engagement:
Engagement levels were notably higher for posts where users had a personal connection with the curator, irrespective of the content type. Contrary to initial expectations, there was a lack of significant interest in discovering new curators, likely due to the limited scope and content of the alpha test.
Several users indicated that they sometimes interacted with posts without actually viewing the video content. This behavior was often due to having previously seen the video (e.g., 'I love this video!'). In other instances, the video's title, thumbnail, or caption conveyed a clear message, making it unnecessary to watch the video to understand its sentiment.
Several users reported experiencing 'post anxiety,' where they hesitated to share content. Some expressed feeling overwhelmed by the sheer volume of available video content, making it difficult to decide what to share. Others indicated that their reluctance stemmed from the platform's current limitation of sharing content with 'everyone,' expressing a desire for the capability to share videos with a select audience, such as friends but not parents.
User Feedback
Kopextacy
"I love being able to find small communities with niche interest and to be guided by what people actually like rather than mindless algorithms. It feels more personal here, in the best way."
SpookyLandrock
"It's really amazing to have a dedicated platform to share fun and exciting videos based on whatever vibe I'm feeling that day. I also really enjoy seeing what other people post and share."
Noel Bass
"I'm a big YouTube video watcher, but YouTube's platform doesn't encourage the social aspect that Malk does. I go to Malk to find content I otherwise wouldn't be aware of on YouTube."
Floyd90
"Malk makes it easy to connect with communities that share my interests. It feels like a more genuine way to discover and share content that really resonates"
Key Learnings (continued)
User Interface:
Feedback on the design, layout, and features was very positive across the board:
- Users found video sharing to be intuitive and fast, and appreciated the ability to post videos from multiple platforms. As a suggestion to streamline the process even further, some users requested the ability to search other video platforms from within Malk.
- Users enjoyed the ability to discover new content via category feeds and tag feeds. Some suggested the addition of a mode that autoplays from one video to another.
- Users loved when people liked their videos or followed their profiles, but wanted to know more about which users 'liked' a post or 'followed' a profile.
- Users expressed interest in features that would enable them to find people in their social circles that were already on Malk, as well as the ability to invite users to the platform*.
- Feedback also included the desire for more ways to share their posts and profiles outside of Malk (ex: Facebook, reddit) in order to more effectively gain followers*.
Performance:
The prototype held up very well for the purposes of the alpha test, and users expressed satisfaction with how fast content loaded and how quickly their posts would show up. However, I found certain aspects of the build that will not hold up at scale:
Filtering Dynamic Data
The primary challenge arises from utilizing the Webflow CMS to dynamically present content on the Webflow front-end. When filtered content is displayed using the Webflow CMS (e.g., only videos from users that a particular user follows), the filtering process occurs post-fetch from the database (Airtable), rather than by directly requesting the filtered content from the database. This results in performance issues related to speed and the accurate loading of content.
Data Limits
Moreover, the Webflow CMS imposes stringent data constraints that limit the volume of data it can accommodate, rendering it unsuitable for a larger beta test and subsequent phases. To address this challenge, I intend to circumvent the Webflow CMS by directly transmitting dynamic data from Airtable to the Webflow frontend, a process I anticipate achieving through the utilization of a platform known as Wized.
Next Steps
Overall, I viewed the alpha test as a great success and validation of the concept. The feedback confirmed that Malk fills a gap in how people discover and share videos, and it also surfaced new directions to explore.
To move forward, I’m focusing on three key areas: infrastructure, feature development, and community growth.
Rebuilding the Backend for Scale
The alpha test revealed limitations in using the Webflow CMS for dynamic content filtering and data volume. To support a broader rollout, I plan to rebuild the backend using Wized to bypass Webflow’s CMS altogether. This will allow me to:
- Query Airtable directly for filtered and personalized content
- Improve page load times and reduce performance bottlenecks
- Eliminate constraints around the number of items the platform can handle
This transition will also make the platform more flexible and scalable for future updates.
Adding Features Based on User Feedback
I’m prioritizing the development of features that users specifically asked for, including:
- Private sharing controls
- Social graph features
- Post and profile sharing tools
- Like and follower visibility
- Autoplay and platform search
These features are aimed at deepening engagement and making the platform feel more social, personal, and easy to use.
Preparing for a Public Beta
Once the backend is rebuilt and core features are in place, I plan to launch a public beta to a broader audience. I’ll be experimenting with onboarding flows, curator discovery tools, and lightweight moderation systems to support a growing user base, as well as observing what kinds of communities and content naturally emerge when Malk is open to anyone.
