This Week:
- The 3 crazy new features I added to my software this week.
- How the second video on my rebranded YouTube channel performed.
- Walking all around Porto trying to find a place to record my next two videos.
January 2025
| Su | M | T | W | Th | F | S |
|---|---|---|---|---|---|---|
| 29 | 30 | 31 | 1/1 | 2 | 3 | 4 |
| 5 | 6 | 7 | 8 | 9 | 10 | 11 |
| 12 | 13 | 14 | 15 | 16 | 17 | 18 |
| 19 | 20 | 21 | 22 | 23 | 24 | 25 |
| 26 | 27 | 28 | 29 | 30 | 31 | 2/1 |
Where In The World Was I?
- 🇵🇹 Porto, Portugal
Metrics From The Week
Business Update
It felt like forever since I actually spent some time building new stuff for my software. This week I got back to it.
New Feature – Viewer Insights
One of the first thoughts I have when I see outlier videos, both on YouTube and on my software’s video explorer page is: why did this outperform? And this was also validated by a conversation I had with a creator, who said that there’s nothing out there that tells you WHY something is an outlier. What resonated with people, emotional or psychologically?
So I uh… built it 🤷♂️
Which shows…
I added a button to the details section of every video. Clicking it will analyze the comments of the video specifically looking for what resonated with people and give you a concise breakdown.
It looks pretty simple and straightforward, but it took me a bit to get the right format for the output. I’m a little skeptical of AI-generated output nowadays and I think others are too, so I made it focus on presenting data, the actual comments, rather than what might seem like “made up” reasons.
One thing I’ll note is that analyzing the comments isn’t a 100% solution since not everyone comments, but I think it’s good enough to use as a representation, or a sample size, for the real reasoning.
I like this feature a lot, I’ve already been using it.
New Feature – Improved Video Ideas
One of the things I was noticing with the video ideas being generated and something I heard from some users as well, was that some of the ideas weren’t relevant. It tended to be a mix and depended on the user, many were really good, some were decent, and some were just not good.
Part of my software’s core value proposition is that the video ideas generated are actually good and personalized, so it was finally time to revisit my MVP algorithm and make it better.
I won’t go into too much detail here because my algorithm is technically top secret highly classified intelligence that only myself and the president have access to under the “Ultra-Elevated Absolute Priority Nuclear” clearance level…
That said, it basically ups the number of ideas generated, then ranks all the ideas based on how similar they are to the user’s channel and how “good” of ideas they are, then picks the best from the batch to show to the user.
And man, the ideas are way better. Being able to get rid of the bottom ~50% makes a big difference.
There’s not much to show for this visually, but here are some of my updated ideas:
New Feature – Your Recommendations
The best feature for last. This idea came up a couple weeks ago as soon as I started actually using my software to get video ideas for my new rebranded channel.
One of the thoughts I had was, so I have all these good videos ideas… But which one do I make next? And why? What should my strategy be?
And a big tell for this problem was that I literally ended up pasting my entire list of video ideas into ChatGPT and had it guide me for what strategy to follow and which idea to pick for my first 3 videos. That’s when the lightbulb went off and I wrote the idea down.
What if I could just tell my software what my goals are, have it look at my past videos, look at what’s working in my niche, analyze comments, and have it recommend a type of next video, then find some examples that fit that type?
So I uh… built this too 🤷♂️
And man the possibility are endless with this. I had so many ideas for what this could do that it was hard to figure out how to just get something started.
I wanted to get the framework laid down first so that I could build upon this algorithm later. So, I started with: have the user enter their goals, their product, who they’re targeting, and if they’re looking for something specific. Then look at the past couple videos, look for gaps and opportunities, come up with a strategy and type of video, then filter through all the video ideas for 3 of the best ideas that fit this strategy.
And back to the AI-skeptical mindset, because I knew people might be skeptical of this (mostly me…), like “why should I believe these recommendations?” I added a modal popup to show the detailed reasoning trail and train of thought specifically based on what you submitted. Then if you’re looking for something different, you can update your prompt and resubmit.
And the best part was that I realized it was right 😂 I’ve been focused on interesting videos that tell my personal story and indirectly share my software, but nothing that’s more focused on my target customer and creating a video to establish authority.
I think this is hella cool. It was so cool to see this working.
When I finally pieced this all together, I had that warm fuzzy satisfaction feeling. There’s just something about having an idea in your brain, figuring out how to actually make it real, then actually bringing it to life. Such a cool feeling and why I love building.
And I don’t want to get too ahead of myself but I think this might be the beginnings of the highest ROI feature and USP (unique selling point) for my software… If it can actually identify things accurately, which is what I’m focused on, then IMO this is massive. I have so many more ideas for this, but ya gotta start somewhere.
I finished this feature and thought “MAN this is not worth $19/mo, I’m selling myself short.”
Personal Brand Update
Day In The Life
This week’s video went live:
And like I mentioned last week, this was fun to edit, but took longer than I thought. I spent a long time syncing up the cuts with the music.
The stats ended up being:
- Planning time: 45 min
- Shooting time: Filming most of the day
- Thumbnail: 1.5 hr
- Editing: 4.5 hr
- Sound: 9.5 hr
Not as much of a smash so far as I had hoped
And I ended up changing the title from “Why I Traded Comfort for Freedom (Day In The Life Working From The Road)” to “Day in the Life of a Van-Life Digital Nomad (Why I traded my comfort for freedom)”
But what I realized looking back at my notes was that this wasn’t designed to be a Hero type of video. It was more about deepening connection rather than attracting new viewers. So with that frame of mind, I’ll take it.
The strategy I’m applying to YouTube and what I’m learning is that not every video is/should be designed to go viral or be designed for attracting new people. I like this methodology, it’s just now about re-framing expectations.
I also don’t want to get too much into the analytics mindset. I briefly mentioned this last week. I just wanted to share the first couple new videos since they’re part of this transition I’m going through.
An Algorithm Peek
Also I think I figured out one part of how the YouTube algorithm works.
If you look at these two last analytics graphs, it looks like what it does is for the first 48 hours, it serves the video to your subscribers (or maybe people that typically watch), then at the 48 hour mark (maybe based on how it performed with that group) it picks it up and starts recommending it to new people on their homepage and through recommended. Kind of interesting!
Onto The Next
And with that, I’m out of van life videos. I only recorded 2, so the next ones are now from backpacking life.
I recorded 2 more videos this week.
Not sure if I’ll use these but taking these thumbnail photos ON the bridge was ridiculous and very awkward 😂 I just needed someone to take a picture of me with the #InfluencersInTheWild caption 😂😂
For recoding, I remembered this beautiful overlook that I decided to go back to. Great views!
From August 2023 (Why You Should Go On The Pub Crawl Twice Instead Of Once August 20th, 2023):
Anyways… I couldn’t get a good angle for recording a video so I went to the back corner of the parking lot under the bridge 😂 🤦♂️ Maybe a questionable director’s decision but let’s see how the video ends up 😂
And since I didn’t want to record my entire video here I walked around Porto for maybe 2 hours trying to find another place I could finish the video. Somewhere: without too many people, level, and shaded. This was more difficult than I thought but I did find a place.
I found a beautiful garden park for my second!
And speaking of personal branding, the waiter I met in one of the coffee shops I went to pointed out my camera and asked I what type of photos/videos I take. He shared that he does a lot of Instagram branding and photoshoots (and they’re very good!). We exchanged Instagrams. Pretty cool!
Finally after 4 years having an active Instagram page has come in handy haha 🙂
Porto Snaps 📸
Town hall square:
Around town:
The bridge/river:
I found out that the hostel offers coffee (espresso), so I moved my morning alarm up to 7:05am (don’t ask 😂; never got around to moving it up another 5 min lol) and started working in the hostel instead of going to a coffee shop first thing.
Social XP
I was up on the bridge and decided to get a social “rep” in asking for a photo of me, which I typically don’t like doing (because it’s uncomfortable).
I went up to this couple, asked, and the woman took my photo.
She was handing me back my phone and asked “is it okay?”
As I was taking the phone, I said “yeah it’s perfect, thank you,” didn’t look at the photo, and immediately put it in my pocket, then walked away 😂😂
The moment she held my phone to take the photo, I had already got what I wanted.
It wasn’t about the actual photo 😂, but here it is:
Food From The Week
Please refer to last week’s pictures 😂
- Honest Greens – bowls
- Callejero – burrito
- Honu Poke and Poke House – poke bowls
- Esquires Coffee – omelette and acai bowl
- Turmeric – indian food
- Nood – pad thai
BUT I did go to 3 new places *crowd gasps in shock*:
- Noshi Coffee – golden latte (turmeric, ginger, and other things I forget) and chicken “noshi toast”
- Cielo Brunch Garden – strawberry smoothie and “healthy brunch” (acai bowl, whole grain toast, scrambled eggs and added smoked salmon)
- Restaurant Pizzaria Rei da Gula Porto – pizza
- Glanni – gelato over waffles
And I’ve been going to Continente for protein shakes and berries (blueberries and raspberries) more as afternoon snacks.
And when I went to Nood this week, I walked in, the server was in the back and motioned for me to sit anywhere I wanted. I sat down in one of the seats I usually pick.
He comes by and goes “pad thai, extra chicken?”
I smiled… “Yep!” 😂😂
What’s Next?
🛫 Packing my bags this week 👀