I'm getting back to regularly writing Build Logs on Sundays, sharing updates on what I'm working on and what I'm finding interesting.
Lessons learned from the first week of launching a YouTube channel
Within the past week, I've posted my first two YouTube videos, so I wanted to share what I've come to learn about the platform and posting so far:
- Retention Graphs: Doing some initial research into being successful on the platform, the two most important metrics seem to be the clickthrough rate, and the retention your videos have (the % of viewers who watch through to a certain point in your video). The goal is, regardless of your retention, for your retention graph to flatten out as quickly as possible, so that after your initial fall-off (there will always be people who click onto a video and click away quickly), the viewers are remaining until the end of the video. Currently the graph for both videos is flattening out at around 30%, which I'm happy with.
- Watch Time: watch time will be important, as it's one of the metrics that determines when you're eligible for the YouTube Partner Programme (the requirement being 4,000 hours of watch time within the past 365 days).
Here are my analytics from the first 5 days of starting the channel.
In a recent interview, Mr. Beast, the platform's largest creator, said that you shouldn't focus on metrics too much before your first 100 videos, so I don't plan to, other than better understanding what is resonating and what isn't.
I've got a lot of plans for the channel in the upcoming months, and achieving them will be dependent on finding systems that work for me to manage the channel as efficiently as possible, alongside other commitments - for now, I'm working without an editor as I want to operate the channel at a profit from the beginning, and it currently isn't monetised.
Shipping every day
Going forward, I'm going to be shipping every day (aside from Sundays) - I've decided this for several reason:
- Increased pace of learning: by running a series of smaller iterative experiments, your pace of learning just becomes so much faster. Instead of building in the dark for months, only to find that your initial hypothesis was incorrect, you can continually update your mental models about what you're working on and who you're building for.
- Better understanding of value creation: by more regularly getting feedback, it's easier to gauge whether there's true value being created in what you're working on
What counts as shipping? Any work that is published live in some capacity will count, including blog posts, code and new videos on my YouTube channel. Not all of these will be featured on this blog, while some of my projects are still in stealth, but it will be happening and I'll share stats on my consistency here.
Laying the foundations for Positive Sum Labs, my personal holding company
In previous build logs, I've written about my intentions to lay the foundations for a personal holding company this year, which I'm calling Positive Sum Labs. Today I've launched a website to keep track of the projects under my personal holding company, and will be updating it regularly.
What's next for Artificial Intelligence?
The final few months of 2022 brought with it a tremendous amount of advances in the applications of artificial intelligence, and it seems like 2023 will be no different. I'm excited to see what the second and third order consequences of these innovations are - it's often the case that it takes people a while to understand the true value of new technology. This blog post by Max Hodak has excited me by helping to imagine a potential future for the applications of AI in the coming months and years.