Visual Brand System With AI: From 64 to 200 Substack Subscribers In 2 Weeks
I Stopped Dressing My Brand for Approval and My Growth Took Off
This post is part 2 of a series on Branding with AI. Read previous posts by clicking here.
TL;DR: I went from 64 to 200 subscribers in two weeks by changing one thing: replacing one-off visuals with a repeatable visual system. I used the SUCCESS Model to evaluate my brand. Most AI visuals fail because they try to score evenly across all six dimensions and end up memorable in none. Pick two or three and dominate them. The free Visual Brand Distinctiveness Test lets you run the same diagnosis on your own Substack in 10 minutes.
I didn’t post more, didn’t change my topic or get lucky with a viral post. I changed one system — and subscriber growth followed.
A few weeks ago, I rebranded AI Meets Girlboss. Since then, I’ve been getting messages about how recognizable the visuals are. How people “know it’s mine” and how they “see it everywhere”. I also saw a very real shift in numbers: I went from 0 to 64 subscribers in my first month, rebranded — and jumped from 64 to 200 subscriber in two weeks.
I used the same voice, the same ideas at the same frequency. But I added a different visual system.
December 7th, 2025 — the day the rebrand went live — is the inflection point on my chart. I’m calling that out because it kills the most common objection upfront: this wasn’t “consistency finally paying off,” an algorithm miracle, or a lucky viral post.
It was recognizability.
Choosing a pink, feminine, fashion-centered design for an AI publication was… well, risky. I made a branding decision to be distinctive and memorable, even if that came with risk. Tech defaults to masculine. That’s why my brand doesn’t and it made me recognizable.
And once I saw that, I couldn’t unsee how invisible the problem had been before. This post is about helping you create a distinct brand using AI, that people will remember.
Hey! I’m Pinkie. I help creators build recognizable Substack brands that drive consistent subscriber growth. I share the exact branding systems, prompts, and workflows I use to transform ideas into distinctive, repeatable output. If you’re new to AI Meets Girlboss, welcome! If you’d like to work with me on your brand, let’s collab! 🩷🦩
In this post, you’ll learn:
why a brand needs to be memorable to be successful
how three standout Substack brands win differently
how I used a single framework to course-correct my own visual brand and why it changed my growth curve
how to evaluate your own visual brand system with AI and start rebranding in 10 minutes
Most Substack Channels Are Great But They Are Invisible
This is the uncomfortable middle phase no one warns you about.
Your content is good. Everything looks… fine. And yet, nothing sticks.
If you scroll Substack regularly, you know exactly what I mean. You see competent, well-balanced AI-generated visuals that are interchangeable enough, you couldn’t tell who made what five minutes later.
If that description stings a little, you’re not behind. You’re just in the most irritating phase of building a brand: the phase where you’re producing good work that doesn’t yet build distinctiveness.
I know this phase well, partly because of how I consume Substack myself.
Why a Visual Brand Needs to Be Memorable to Drive Growth
In a conversation recently with Anna Levitt , I realized something. Some brains — mine included — don’t process text first. We process visuals first. The visuals decide whether we click, whether we stay, and whether we ever engage deeply with a creator’s work at all.
“If your brand is text-only, you’re excluding half your audience.”
For visual-first readers, visuals aren’t decoration. They’re the door. And if your door looks like everyone else’s, people keep walking.
What Actually Changed (And What Didn’t)
Let’s be precise, because this is where most growth stories get hand-wavy.
What did not change on my channel:
My topic
My audience
My posting frequency
My writing style
What changed:
I replaced one-off visuals with a repeatable visual system.
A system.
Same character.
Same palette.
Same compositional logic.
Same visual grammar across posts.
That constraint — not creativity — is what unlocked growth.
This is where most creators misdiagnose the problem. They think they need to be more creative or they need more variation. What they actually need is a recognisable silhouette.
Fashion Week doesn’t work because every look is different. It works because you can recognize a house from across the room.
The Framework I Use to Measure Visual Brand Recognisability
When I work on brand storytelling — whether for clients, advocacy programs, or my own work — I don’t start with style. I start with memorability.
The framework I use is the SUCCESS Model, originally from Made to Stick by Chip Heath and Dan Heath, and it breaks down why some ideas stay with us while others slide past. I don’t use it as a creativity checklist. I use it as a diagnostic.
In visual brand storytelling, SUCCESS becomes a recognizability stress test:
Simple – One dominant idea per visual.
Unexpected – A clear break from category norms.
Concrete – Specific symbols, not abstract vibes.
Credible – Signals lived experience or authority.
Emotional – A legible feeling within seconds.
Story – A moment of tension, change, or decision.
Most AI-generated visuals don’t fail because they’re bad. They fail because they try to score evenly across all six — and end up dominating none.
Balance is the enemy of recognition.
Three Substack Brands With Strong Visual Brand Systems
Write10x (now Unpromptable) by James Presbitero
Editorial note: Between drafting this post and hitting publish, James rebranded Write10x to Unpromptable. I’m using the Write10x era visuals here purposefully - not because the new brand isn’t worth studying (it is), but because this moment proves the point of this post: distinct brands can evolve too!
Write10x by James Presbitero is instantly recognisable. James’ visuals don’t aim for relatability; they aim for archetype. I remember his thumbnails for days, and often catch myself reverse-engineering the thinking behind a visual choice. They don’t just attract attention — they invite interpretation. Like a signature silhouette you recognise without seeing the label.
Where it shines:
Unexpected: High-contrast black visuals that polarise and stand out in-feed.
Emotional: Hero imagery charged with ambition and authority.
Credible: Guest authors appear inside the visuals, lending instant legitimacy.
Co-Write with AI by Nick Quick
Co-Write with AI by Nick Quick immediately stands out because it refuses the default AI aesthetic. I found this channel while actively looking for a distinct visual system, and it stopped me mid-scroll. The hand-drawn, almost zine-like illustrations are opinionated, slightly messy and trust the reader to meet them halfway.
Where it shines:
Unexpected: The raw, illustrated style breaks from the polished AI-photo norm on Substack.
Simple: One dominant idea per visual, no visual over-negotiation.
Story: Diagrams and drawings imply tension and cause-effect, not decoration.
AI, But Make It Intimate by Calder Quinn & Kristina Bogović
AI, But Make It Intimate by Kristina Bogović and Calder Quinn doesn’t just have visuals — it has a world. The characters, lighting, and recurring settings pull you in immediately. You’re entering a narrative universe that feels cinematic, emotionally charged, and meticulously maintained.
Where it shines:
Story: Every image implies relationship dynamics, tension, or emotional stakes.
Emotional: Desire, discomfort, curiosity — the feeling is legible within seconds.
Credible: The commitment to character and world-building signals seriousness, not novelty.
Strong Visual Brands Choose One Thing to Dominate
None of them try to do everything at once. Each one chooses what to dominate.
That’s the part most creators miss when they use AI for visuals. AI makes it easy to generate options — which tempts us into balance instead of judgment.
SUCCESS isn’t about creativity.
It’s about recognizability.
Running the SUCCESS Test on My Own Visual Brand
Before the rebrand, my visuals were safe and invisible.
I created a Distinctiveness Test based on the SUCCESS Model to diagnose visual brands with AI. I am sharing it with you at the end of the poot, so you can evaluate your own visual brand and start rebranding in 10 minutes using AI. But before sharing it, I want to show you what happened when I ran it on my own Substack in early December.
The verdict was… humbling.
To be honest, another creator could match this vibe in one Midjourney prompt. Your SUCCESS Score:
Simple – 3/5 A dominant idea exists, but each image carries competing cues.
Unexpected – 2/5 Nothing breaks expectation. They’re polished, but predictable.
Concrete – 3/5 Work scenes, but nothing that differentiates the work itself.
Credible – 4/5 Strong lived professional context.
Emotional – 3/5 All images sit in the same emotional register. No risk.
Story – 3/5 Each image implies a moment, but not a change.
Ouch. My only real strength was credibility. Everything else was beige. As I sat over my laptop with my Substack main page open, I noticed the invisible rule I’d been obeying:
“If I’m teaching AI, I should look neutral, credible, and professional first.”
It got me thinking: that rule protected me but it also limited me. AI-generated corporate women images were consistent, but they were everywhere on Substack. So I asked different questions.
What if I did something unapologetically feminine?
What if I leaned fully into the female gaze?
What if I used symbolism women actually recognize?
That’s how the fashion illustration direction was born.
After the rebrand, I made a deliberate decision to dominate three dimensions:
Concrete: Locked in a visually specific world. Every post cover feels scene-based, even when surreal.
Unexpected: Fashion-editorial illustration + AI education breaks the default mental model of “AI = sterile or masculine.”
Story: Each image implies a narrative: how I experiment, decide, fail, discard, and own my process.
I didn’t get it right immediately. Visual consistency was hard to nail (Note: I’m writing a post on this later in the Branding with AI series). My early covers were cluttered with clever props and inside jokes. I overestimated how much time people spend on a cover image. (They didn’t even notice I grew a third leg on my 100-subscriber post.)
But once I committed fully to the direction, I defended it ruthlessly.
Since then, I’ve been getting messages about how recognizable the visuals are. How people “know it’s mine.” How they “see it everywhere.”
And I saw it in the numbers too:
0 → 64 subscribers in my first month.
Rebrand.
64 → 200 subscribers in two weeks.
I walk you through my full rebrand process step by step in I Let Nano Banana Redesign My Face post, if you want the behind-the-scenes.
Your 10-Minute Visual Brand Distinctiveness Test (Now as a Custom GPT)
Even though I love 1:1 deep dives — like the visual brand session I did with Dallas Payne over the holidays — you don’t need my eye to start diagnosing your own storytelling.
The Distinctiveness Test is now a custom GPT you can run anytime, solo, in ten minutes, and it’s 100% free.
You upload:
A screenshot or PDF of your Substack homepage (grid view)
Optionally: your topic, audience, and publication URL (context only)
It gives you:
A clean SUCCESS scorecard
Your strongest dimensions (and what they signal to readers)
The invisible rule you’re playing by
One focused risk worth testing next
Concrete constraints to improve recognizability
If you want to go deeper, it can also:
Identify your strongest visual and explain why it works
Flag your weakest cover
Help you redesign that cover based on the actual post content
In the comments…
If you try the test, share one of these: Your weakest cover vs. the reworked version or your SUCCESS scorecard, with the one number you’re trying to move
Your Brand Needs a Stronger Point of View
If your visuals are interchangeable, you need a stronger point of view, and you need to take a stance. AI can generate endlessly, but it needs your judgment and you still have to make the final call.
And if you’re a visual-first reader like me, this matters because it’s how a lot of us find each other on Substack.
Talk soon Girlbosses,
Pinkie🩷🦩
Substack University 🎓
In every post, I credit the systems that shaped my thinking. Substack really is a kind of university. Late-night learning. Public experiments. Shared notes. This post was influenced by the creators below. I encourage you to check them out!
Karo (Product with Attitude) is launching a series where she features builders. In 2026 this will be my core inspirational source: Build With Attitude ✊ - An Open Invitation To Vibecoders Who Care About Building Things That Matter
Dallas Payne is in the process of developing her visual style for her publication The Daring Next. Check it out and I recommend reading her hilarious, self-reflective post on the process: The AI tool that works best? Humans.
Jenny Ouyang’s visual system thinking made my rebrand possible in the first place: How I Create Consistent Hero Images, And Why I Haven’t Switched to NanoBanana
Dheeraj Sharma’s post on training Claude to reflect your brand voice is for everyone writing with AI: How to Train Claude to Write in Your Exact Voice?
What you’ll walk away with
A clear read on why your current visuals are (or aren’t) landing
Your brand archetype — the psychological frame your visuals should reinforce
3 concrete visual direction(s) delivered straight to your inbox. You can hand to a designer or execute yourself with AI
A prioritized next-step plan so you stop second-guessing every design choice
What a great idea! That would mean each strand has its unique visual messaging. I'd love to see how it turns out!! Please do tag me when you are happy to share the result or the progress! 👀🩷🦩
I’m using your guide to rebrand this month. I think Substack needed a visual brand designer like you! I consider myself very lucky to have met your work, pinkie! 🩷