Day 143:
Earlier today I was honestly feeling a bit lost.
Not in doubting myself and what I’ve learned so far. I’m still solid in my own conviction.
But I felt I’ve been listening to too many people—like is common in the very beginning—and I started to confuse myself again.
So I’m gonna minimize my inputs and info sources again.
Today I went back to basics.
I brushed up on market awareness/sophistication levels and the product research modules on Origins.
You can probably tell I’ve been exhausted, and I’m realizing this too late but honestly it’s really important that I keep my health up. Basic stuff like sleeping enough, regular meals, exercise a few times a week.
Otherwise I can’t think properly.
So not my greatest day today, but we keep going.
16
465
Day 142:
1 sale today.
I paused the CBO. Does that mean we cut the supplement?
I’d say yes, for now. We can turn the CBO back on in a few weeks and performance would ramp back up within a week.
We’re not moving on because of Black Friday. The holidays are damn near irrelevant for us because we don’t have anything working.
The product-market fit for this supp is validated. It’s just my skill issue.
On the Facebook groups, they are aware of our product and discuss it a lot. The sophistication level is much higher, more skepticism, and a lot of people complain about our comp’s product and customer service.
But I found another product from one of the comments in this group.
My idea is:
- selling to the same + adjacent mkts
- switching from a supp to a otp high aov ($65+) product, and if its giftable, even better
I think the market we’ve been selling to is solid, there’s a lot of consistent buying intent.
And hey, maybe the supplement could be an upsell or bundle item.
I can still look for other products but I want to get faster and get a new test up asap.
this is the most awake I’ve looked in weeks 😂
also training grokbot to edit my vids for me
time better spent researching or making ads
1
14
845
Day 141:
1 sale today.
The last 4 days have been 6, 1, 5, 1. It’s a good real example of what inconsistency looks like at low budget.
All I can do rn is test more ads and find one that does a 180 on the current situation in the ad account.
I launched native and video batches in the CBO tonight and duped the videos into the ABO to get data faster.
The ABO isn’t necessary because if the videos do perform, even if they were only in the CBO,
I noticed good ad sets “float to the top” and get spend over the next few days until they’re consuming most of the budget.
That’s regardless of when they were added—so long as that ad set is performing.
I hope we get sales from the new ads tomorrow.
2
12
727
Day 140:
5 sales today.
Today I trenched so hard. The amount of learning that’s happening right now is immense and overwhelming. I feel there’s so much I have to learn and do.
It’s a lot of hours alone.
It’s a lot of work no one sees.
And speaking for myself, I do think ecom is harder than it looks. At least at the start.
First half of my day I worked on copy/scriptwriting.
Later, I practiced making my own videos.
Copy-wise, I felt like something clicked for me today.
I assembled copy for video scripts, as in moving sections around and seeing how that affects the sales message.
I reviewed some old critique someone gave me, and today felt like I finally understood the script structure Nathan Perdriau talks about in his 3 hour copywriting masterclass on YT.
I definitely need to spend way more time on the research side.
I already felt like today was a huge unlock but I am struggling with not being organized—I have learnings and notes everywhere—and it’s definitely slowing me down a bit and making my brain messy so I need to take time to reset.
Day 139:
1 sale today.
I wrote 3 video scripts with 5 hook variations on each.
I’m editing at least one of them myself. My automation friend said it’s still very important that I know how I like to produce video ads. Especially because they can be used as reference material to inform Claude’s SOPs later.
You learn a ton by making videos manually. And even when I comment revisions to my editor, it forces me to think about how the frames give life to the script and if I’m hitting the emotions I want to target.
Also sent revisions to my editor. I think we’ll be able to launch his ads tomorrow, which I’m excited for.
Day 138:
6 sales today.
My friend doing consistent 10k days walked me through Claude automation today and our call genuinely lasted 9 hours.
We were productive from 10am to 7pm—sounds crazy but it’s true.
I got set up with the necessary software, then he showed me how to create workflows, on my own, for static and video ad generation (both generating and editing frames together).
We also started establishing SOPs for the ad making workflows.
I practiced making statics, then we made an AI slop song ad just to run through the video ad making process once.
Obviously I have to apply direct response principles to my work, otherwise I’ll just produce AI slop faster.
I still have to make ads manually for now because it does take a lot of time and work to get Claude dialed in per your SOPs
But once you do, it’s definitely a game changer. It’s incredibly high leverage.
He gave me homework to complete this week so I can actually practice the automation work.
Rest of today I launched a new batch.
I got ads back from my editors but they need to go through revisions again, so there’s still a ton of manual work but the skills all compound and will carry over to when I dial in automations.
7
41
1,777
Day 137:
3 sales today.
I made a lot of video ads today manually. Tried 2 different styles: realistic UGC and Pixar-style animation.
Practiced writing hooks—script and visual.
And I sent revisions to my editors so I’m waiting for their ads.
I have the Claude automation call with my friend tomorrow, so I signed up for some softwares and prepped different styles videos to kickstart the workflows.
Day 136:
1 sale today.
Today I talked to a friend about how he uses Claude. All of his static and video ad production— meaning clip generation AND editing—is automated.
He just focuses on writing solid scripts with direct response principles.
For me, the technical ad making process is what’s wasting a lot of time. Time that could be spent learning marketing.
One of my main problems is speed. My ad output is lower than I want because it genuinely takes me a long time to make them. And that’s when I realized I’m doing something very wrong.
We’re gonna call on Sunday and set up some workflows. I’m super grateful.
Then my job is to build out and refine the workflows for our product and market.
My goal is to set up a solid enough automation process so I can redirect focus to learning marketing and creative strategy.
Gotta keep my mind right. I have to focus more on the learning process than the actual results we’re getting, otherwise I get sad.
10
23
1,551
Day 135:
2 sales today.
Today I trenched video ads and learned a lot.
Editors got back to me with 3 different videos but I need to make revisions on all of them.
I was told there’s a more efficient way to automate editing so I reached out to 2 people to learn how they do it.
Watched Nathan Perdriau and MBB’s videos on hooks. Then I wrote 5 different hooks, both script and visual hooks, for 4 different videos. Total of 20 unique video ads we’ll test across 4 ad sets.
Also called with my friend today regarding video ads. I learned so much from him today, especially re how he generates natural-looking AI clips.
Big learning day but I’m still:
1. Slow
2. Inefficient
3. Technically retarded
Next I need to learn what Claude can handle and is worth my time/energy investment.
4
12
813
Day 134:
2 sales today.
Today I finally learned how to work with video editors.
I had been putting it off and felt resistance toward it because it was something new.
But we need video ads so I dove into it today.
Each video has 5 different hooks so I can test 5 different videos in an ad set.
I gave them a reference video and frame-by-frame instructions because I wrote the scripts myself. ChatGPT came up with clip ideas for me.
I don’t think this is the most efficient way to do guide them, and its slower for me to put together instructions. And I’m not sure my script structure will perform because I wrote it myself.
So next time what I’ll do is swipe a scaled video’s structure, then have them replace the mechanism, product, and offer.
That should move faster because they can reference the proven video for structure and and style.
Definitely learned a lot today! So much to do.
7
18
1,064
Day 133:
1 sale so far today.
We’re still running the supplement.
I wrote a new script off our best performing native and having editors make videos for us.
After going through the Origins product research content, I realized how important it is for us to differentiate from comp.
You can differentiate in 2 ways:
1. Product/mechanism
2. Avatar
I’m gonna try to differentiate on the avatar. I’m working on a new PDP/statics ad set for a different but adjacent avatar to the one we’re targeting.
None of the comp speak directly to this avatar and I want to test it.
The concepts that are showing promise for us are also different that the ones our comp are scaling with. Not sure what that means.
I’m gonna go through the avatar modules in Origins and apply them to this supplement/market bc I feel we’re lacking on the research side
2
14
733
Day 132:
Made 2 sales today.
I finished the product research videos in Origins and have a few product ideas.
Partner and I are gonna pick one and build out the PDP and ads tomorrow.
I’m focusing more on speed, specifically testing ads daily/every other day.
It’s not mandatory but if we can get a product working for Black Friday, that would be ideal.
But I’m generally going for evergreen problem solvers that we can sell year round.
4
18
1,090
Day 131:
Made 3 sales, ~$200 in revenue today, from the new static batch added yesterday.
But I’m still looking for a new product.
Today was Sunday, reset day for the week ahead. I was thinking about things I need to do differently/better because my current approach needs improvement.
The goal for next week is locking in on a product and avatar research.
Today I bought a new notebook so I can take notes as I go through Origins.
I also watched some of the product research module and next week after I find a product I’m gonna dive into the avatar module.
I’m definitely relearning ecom in a more structured way.
I started playing with Grokbot per my partner’s rec, but I also had another friend put me onto automating with Claude automations. Not just for statics but also to handle various workflows.
So I’m trying both, not at the same time, but seeing whether I like Grokbot or Claude more.
4
14
895
Day 130:
Today I launched static batches of our best performing concept.
We’re giving it a day or 2, but I’m close to moving on if nothing from this ad set hits.
Also wrote a 1min video script for that concept but I’ve already started looking for a new product.
I’m horrifically slow and inefficient when I make ads because I’ve been prompting manually, so I’ve been curious about automating with MCPs to pump out statics faster.
My partner told me about GrokBot today and I’m definitely gonna play around with making bots.
If anyone has recs please let me know because I am grossly underleveraging AI rn—to my detriment.
7
25
1,906



