Reflections on 1 Year of (Trying To) Become Successful on YouTube

January 10th, 2025

In 2024, my friend and I started making YouTube videos. This article is just some housekeeping I’m doing to keep track of progress for this channel. I hope that some people find it valuable. I offer some advice in this article, but everything should be taken with a grain of salt, as we are not (hopefully yet) a successful channel. If you’re interested in getting in touch, there is contact information at the end of the article.

Who are we?

We are just a couple of friends trying to have fun and create content that people enjoy. More specifically, though, Luke is the only one of us who does this full-time. After quitting his job at Apple (and then briefly unquitting to work at MrBeast before quitting again), he decided he wanted to stop creating so much content for others and really give solo video development a try. He is the face of the channel and undoubtedly the one who works the hardest to make it successful. As for me, I am a software engineer and amateur data scientist who enjoys making and doing cool stuff.

The videos we have made so far

Now let’s get down to it. In 2024, we filmed and released eight videos (we also filmed two more, but they didn’t come out before the new year). Let’s break them down one by one and discuss the production process and what we learned along the way. Each section also includes a link to the video on YouTube if you’re interested in checking it out. I’ll stay away from analytics in this section and focus on the inspiration, production, and my reflections looking back. I’ll cover the analytics in another section.

I Went Skydiving In Apple Vision Pro!

At the start of the year, Luke was working full-time at MrBeast, and it wasn’t until he quit in March that we went all-in on this channel. I’m not sure exactly who came up with the idea (we have a much more formalized process for creating videos now), but as fans of doing crazy/stupid things and as two Apple nerds, it was probably only a matter of time before something like this came to fruition.

For many reasons, we were initially skeptical about pulling this off. There seemed to be so many barriers in our way. Eventually, though, we got in touch with a great company called Skydive Orange, and they seemed at least willing to hear us out. After many emails and phone calls, we finally booked a day to film. At the company, maybe two people had spoken to us, so showing up with our silly little headset and trying to find the right people to talk to was super awkward. Eventually, after a few hours of waiting around a random aircraft hangar in Virginia, we both jumped out of a plane—one of us with a $4,000 computer strapped to our face. Let me stop writing for a second and ask Luke what it was like:

I mean, yeah, it was pretty crazy.

After this, we decided that we wanted to pad the video with a few other fun activities while wearing the headset so that we could build up to skydiving for the finale. We landed on:

1. Basketball
2. Ice Skating
3. A Ropes Course (somebody that worked here gave me their email and asked if I could send some of the footage we got - I immediately lost this card and if you’re reading this I am very sorry)
4. Dinner at a restaurant
5. Biking
6. Watching a sunset (and comparing it to the sunet enviroment inside the headset)

Overall, the video is still one of my favorites we’ve made. It didn’t perform quite as well as I hoped, but looking back, it could have gone way worse.

I Hiked A Mountain Blindfolded!

This video was the one I had the least involvement in. It was largely organized and planned by Luke and our other friend Siah. Unfortunately for all of you, Siah isn’t writing this article, but I’ll do my best to provide an accurate recap anyway.

The idea was that skydiving in a VR headset (Apple, don’t shoot me for calling it that) might be a bit too niche, and we wanted to broaden our audience. Filming this was rough, and Luke’s ankles definitely took a beating as he and Siah raced to reach a specific overlook by sunset. We spent a lot of time beforehand brainstorming ways to make the video interesting along the way. This was an example of a video where we felt the concept was intriguing, but we struggled to figure out how to turn it into a full-length video. Often, the inverse would happen as well.

I Let AI Control My Vacation!

Looking back, this is my least favorite video, but it might have been my favorite to make. I did undergraduate research on NLP and had built some rudimentary language models during my collegiate career. All this to say, I’ve been a bit annoyed with all the language model hype and have purposely strayed away from it in conversation. I feel like this article characterizes my feelings well.

I’m not sure who came up with the idea, but Luke pitched it to me, and although I wasn’t entirely on board, it sounded like we were going to have fun, and in that case, I’m always game. We ended up doing a pretty classic trip to Orlando, Florida, and I think we both slightly regret not ensuring something crazier happened in the video. That being said, this was a ton of fun to film, and in some ways, I was happy when it flopped, as it meant this would be the last time we made a video like this.

Find Me, Win $1000!

This is a frustrating video to talk about. For obvious reasons, it was one of our more costly endeavors, and it didn’t go according to plan. Generally, when we come up with video ideas, we try to avoid the game show-esque side of YouTube, but we thought this concept was pretty interesting and wanted to put a fun spin on it. Both of us grew up in Washington, D.C., and are still based in the area, so naturally, we planned a city-wide scavenger hunt that would lead players to Luke, $1,000 in hand. We planned clues, locations, routes, fun costumes, transitions, and created graphics for the live stream. I even built a custom web application to track player locations and display a closing circle that updated every twenty minutes, narrowing down the search space. We did all this, only for Luke to be found in under ten minutes by three different people.

We were not expecting as much participation as we got and were worried about nobody playing at all, but to our surprise, there was a pretty good turnout of people both playing and watching the stream. We posted the highlights of the prep and main event on YouTube, and it performed about as well as we expected, doing slightly better than all of our previous videos.

One of the frustrating things about this is that, coincidentally, some other people had similar ideas soon after we released this video, which in turn inspired a bunch of other variations. We still think we did it the best and would love to make a sequel—we’re just actively waiting for the funds to fall into our lap since we’d like the prize to be larger.

I Spent A Day In 3rd Person!

This video has had quite a comeback. It originally started with the title Day In The Life Of A Drone and was supposed to revolve around Luke flying a drone, attempting to complete everyday tasks (somehow, we thought that was a really good idea at the time). Eventually, it morphed into its current iteration when we realized the idea of doing things from the perspective of a flying camera was way cooler. Initially, the video performed poorly but gained traction after a portion of it was posted by Daily Dose of Gaming.

One of our only regrets with this video is not adding a little indicator on the screen every time a shot was from the drone. We thought it was pretty clear during the editing process, but some initial feedback after release suggested we may have been overly optimistic.

A Conversation With My 20 Year Old Self

Including this is questionable. This is a recurring series Luke is doing that exists outside of the JCU (LukeJ Cinematic Universe). I won’t give a full breakdown of it, but definitely check it out—it’s a cool series!

GeoGuessr In Real Life - Blindfolded & Standed In A Random Location!

This is our most successful video to date. In many respects, we’re often just trying to recreate it in any way we can. We’ve come to think there are three different criteria we want to hit with each video:

1. It should be fun to make.

2. It should be fun to watch.

3. It should be fun for our wallets (affordable).

We felt like this video really hit all three. I chose one of the cheapest flights I could find, Luke and I almost always enjoy traveling (especially for a video), and it performed significantly better than any other video we’ve made thus far. We felt like we took a concept that had been done before and improved on it; I don’t believe anyone has actually been flown somewhere blindfolded for a GeoGuessr YouTube video (sorry to the lady scanning our boarding passes—she was definitely very confused by Luke’s situation).

The video was also relatively quick to make—filming only took a single day, and we got to explore Providence, Rhode Island, which neither of us had visited before.

Overnight In The World’s Most Haunted Airbnb

We wanted to make a Halloween video and thought Airbnb videos were still pretty trendy. Neither of us believes in ghosts, but we thought it would be fun to open our minds for a bit and create a lighthearted ghost-themed video. We were partially inspired by a similar video that we really liked, but we wanted to put our own spin on it. A fun part of this project was that I got to experiment with electronics and build a battery-powered temperature, infrared, and electromagnetic field sensor.

Overall, we were disappointed with the performance of this video. Looking back, it’s unclear exactly what went wrong, but we suspect the video may have been restricted due to some unfavorable policy around the semantics of our title and content combination. I discuss our discoveries about restricted videos in a later section.

Analytics

Now, let’s talk about numbers. When I started writing this section, I think I drastically underestimated how much I would want to write down, so I’ll save most of it for its own article. I think video analytics (specifically on YouTube) is something that’s really easy to obsess over, and there’s tons of content out there about how you can optimize them. The problem with a lot of this content is that it’s mostly anecdotal or overrepresentative, focusing on individual experiences or specific cases in which certain factors helped a video succeed or caused its performance to tank. I’m sure there’s often a grain of truth to this type of content, but for anyone who’s spent a considerable amount of time messing with YouTube’s public APIs, RSS feeds, or just observing YouTube Studio frequently after uploading multiple videos, it becomes clear that predicting video performance, especially for a small channel, is a function with a super high dimensionality. A lot of time can be wasted trying to game the system without truly understanding what you’re getting yourself into. If I could offer some advice to people starting out, though, it would be to focus on making content you enjoy and are passionate about, and just keep getting better at making it.

Ignoring how poorly our most recent video performed, I’d like to think we’re moving in a somewhat positive direction.

How we make a video

We break down the production process into a few stages.

Hopefully, the graphic provides a relatively adequate explanation for each step. We are both very involved with every stage, except Post-Production, where I lock Luke in a cave and tell him he can’t come out until the video is fully edited (this is actually what he’s doing right now while I write this article).

How we decide what’s next

This process has fluctuated quite a bit. We started by mostly just word dumping a bunch of ideas into a shared note and then whittling them down to a select few. We noticed, though, that this led to a bunch of ideas we weren’t super excited about, and most of our best ideas came about when we were just hanging out and casually talking about videos. Trying to rely solely on the latter, however, meant we weren’t generating good ideas fast enough. We recently settled on something we call Title Tournament. Once a week, we both pitch eight video ideas and put them through a bracket, where we vote on the winners of each round. This method has allowed us to have fun with the ideation process while also raising the standard of our pitches.

We also use various tools to help out; a lot of ideas come from just browsing and discussing other videos on the platform that we like and figuring out how we can put our own spin on them. All this to say, we have no shortage of ideas we’re pursuing at the moment, and we hope that the next year will allow us to produce a ton of high-quality videos.

Mistakes we made allong the way

Restricted Mode

So, YouTube has this feature called Restricted Mode (you can enable it in your settings). It’s something that I think both Luke and I subconsciously were aware of but didn’t put a lot of thought into since we were more focused on hidden restrictions that YouTube doesn’t explicitly tell people about. It wasn’t until about two months ago that we realized that when you turn on restricted mode, the following videos disappear:

• I Hiked A Mountain Blindfolded!

• I Let AI Control My Vacation!

• I Spent A Day In 3rd Person!

• Overnight In The World’s Most Haunted Airbnb

You know, just our worst-performing videos by far. We have since been able to get them unrestricted, but this was quite a shock/revelation for sure. It may just be coincidence, since there are various channels that perform well despite having the majority of their videos restricted, but this was still a crazy realization nonetheless.

Decision Paralysis

It’s super easy to get stuck in the cycle of spending many hours, which ends up feeling like wasted time, trying to come up with a new video idea. This is often followed by getting so tired of ideation that we just pick the best of whatever we have so far, only to eventually quit that idea because neither of us were ever passionate about it. Rinse & repeat. Similarly, when planning or filming a video, often unexpected things would happen that would leave us questioning where to go from there and how this was ever going to work in the video’s story. For example, when Luke was found in Find Me, Win $1000! after only a few minutes, it had us wondering if we were just going to have to make that whole video a short.

All this to say that it’s easy to get stuck in frustration, as any part of the video-making process may throw you outside of your comfort zone and push you away from the content you really want to make. I think the usual answer to decision paralysis is to just pick something and run with it. There will be time later to change your mind or re-decide with your newfound knowledge. Although I don’t totally disagree, I think content creation often works a little differently. Sure, you can spend ten years making every mistake in the book and finally land on what works for you, but I also think we’ve benefited a lot from just refining our processes. The more structure and protocol we provide to the content generation process, the more predictive we are of things that may go wrong, and therefore the less time we have to spend thinking about how we want to adapt.

Conclusion

It’s been a crazy year, and although we haven’t had the success we would’ve hoped to get straight out of the gate, I think we both feel optimistic for 2025. If you are interested in chatting, collaborating, or sponsoring us, feel free to reach out to me at [email protected].