If I were starting a brand from zero, I’d build the system first
How I turned taste into infrastructure after one very public mistake
I remember staring at the LinkedIn ads dashboard, scrolling slowly.
There’s a specific kind of dread when you realize something is wrong but you don’t yet know how wrong.
The ads were live. They had been live.
And they were off.
Not dramatically off. Not catastrophic. Nothing that would make headlines.
Just… wrong.
A capitalized word that shouldn’t have been. A phrase our buyers would never use. A pluralization that subtly signaled we weren’t insiders. A design that felt subtly like a different company.
In any industry, there’s insider language.
If you mess up one little thing, pluralizing a word they wouldn’t, getting the capitalization wrong, using a word they would never use, these are seemingly small things that can create a world of separation between you and your buyer.
You can go from insider to outsider in a second.
How did this happen? Probably we hired a new contract designer and someone was trying to move fast and shipped something that wasn’t quite right.
And I had missed it.
Then I got a nasty message from someone on our team. Or, I didn’t get it. They posted it in a public Slack channel full of way too many people, just railing on a set of small mistakes that had gone out and were now running live as ads.
My chest went tight. I couldn’t even feel defensive.
Unfortunately, that person was right.
(But…putting everyone on blast in a public space is not necessarily what I would recommend either.)
This had slipped through without me noticing, meaning I didn’t have the right process in place to ensure this didn’t happen. When I wrote about what I’d do differently in my first 90 days, this was one of the blind spots.
But it could have been worse. It could have been an influencer or, god forbid, a customer railing on us in a public Reddit thread with thousands watching. Beating up these small mistakes in a public Slack channel with maybe a couple hundred internal only folks? Ok, could have been worse.
Was still a hard day.
I probably went home and made spaghetti and ate it in my bed before curling up into a ball.
I don’t remember all the details of that campaign and how it got created. Probably blocked them from my memory from the embarrassment. But I do remember feeling like an absolute failure and wanting to crawl into a hole and never do anything creative again.
Brand was living in my head
For a long time, I carried brand as intuition.
I knew what sounded right. I could feel when something was off. I could spot insider language instantly. I was the keeper of the brand.
That’s taste. And taste matters.
But taste that isn’t documented becomes fragile. As soon as the team grows, as soon as contractors enter, as soon as you increase velocity, taste becomes tribal knowledge.
And tribal knowledge eventually slips.
That day taught me some important things.
Lesson 1: We reinvent the wheel because we’re bored
First of all, sometimes we go off script as marketers because we’re bored. But some of the best brands out there are still working the same messages that worked for them for like twenty-plus years.
I remember when I was at the Exit Five Drive B2B marketing conference in Vermont last year, Sangram Vajre in his talk said something like:
“Get really good at something and just be boring at it.”
The best brands don’t chase novelty. They repeat what works until it compounds.
That ad campaign wasn’t a creative failure. It was a systems failure.
Lesson 2: Brand can’t be tribal knowledge
Second of all, we need a really tight system for making sure it’s going to land with the buyer and avoid small mistakes.
We have too much technology at our fingertips to make these kinds of mistakes. And it doesn’t need to be a whole set of resources to review every single word.
Take the time to document the rules, the principles.
Are there words we never use? If those aren’t written down somewhere, why are you making this harder for yourself?
This should not be tribal knowledge.
One of the things I love about working with AI actually is that it forces you to think systematically. You need to write down all the rules, the learnings, the steps.
Lesson 3: Reinventing from scratch is slow
Thirdly, that set of ads was a pain to create actually.
All net new approach, every new component and word agonized over by someone on the team.
We must not reinvent the wheel unless we have to. And ironically, you can move way faster if you don’t reinvent the core components.
If you’re putting together pre-built pieces versus creating everything from a blank slate every time, it helps you move way faster.
Reinventing from scratch feels creative. But it’s slow. And it increases the surface area for mistakes.
Brand, messaging, and design should feel more like putting Lego blocks together than starting from scratch.
Not approaching brand as a scalable system is risky and time-consuming.
When brand lives as a scalable system, you don’t eliminate creativity. You focus it.
How to build a modular brand system
After that campaign, I built infrastructure. Not a PDF brand guide that no one reads. A working system.
Part 1: Messaging
I have a process where I create a bunch of messaging building blocks. These can all be put together in probably many different AI writing setups. Lately I’ve just been throwing them into a Claude project.
The Manifesto — the reason why you exist:
Before messaging pillars, before format playbooks, before component libraries, there’s the manifesto.
I’ve written an entire piece on why the first thing I’d build as a first marketer is a manifesto. It’s not a homepage headline. It’s a conviction document. The belief system that informs everything downstream.
Once that foundation is clear, you can build the operational layers.
Brand Brain — the strategic foundation. Includes:
Positioning statement
Elevator pitches (one-liner through full paragraph, copy-paste ready)
Value propositions (overall + by audience segment)
Key messages / pillars (each with supporting proof points)
Buyer journey messaging (awareness, consideration, decision)
Product messaging hierarchy
CTA framework
Approved proof points (stats table with usage rules)
Voice & Tone — how the brand actually sounds. Includes:
Who we sound like
Core voice principles
Writing guidelines
Do’s
Don’ts
Words and phrases we own
Words and phrases to avoid
Quality check
Audience Language i.e. IYKYK Glossary (optional but cool) — insider knowledge about the specific audience so that everything you write understands their world from the inside. Includes:
Insider glossary (terminology with context for why each term is loaded)
Example messaging per term / how to use it
Guidance on how to signal “we’re one of you”
Format Playbook — rules for every content type you produce (emails, LinkedIn posts, case studies, etc.). Each format entry defines:
When to use it
Required structure and section order
Length guidelines
Headline and subject line rules
Author / sender defaults
Format-specific do’s and don’ts
Voice and tone adjustments
Quotes Bank — a library of real customer quotes, tagged and scored so you (or AI) know which ones to pull for which situations. Each quote includes:
Customer name, job title, and firm
Theme category
Pain point it addresses
Strength score (1 to 5)
When you feed this into AI, it doesn’t hallucinate a brand that sounds soulless. It operates with a personality (and guardrails).
Sometimes AI isn’t the risk. Ambiguity is.
Part 2: Design
The key with design is building a component-based system.
Start with templates and components
The easiest example of this is on a website. Each webpage type is a template (this is a feature page). Within each page you have components (this is what customer quotes look like).
But this extends to your whole brand.
Think systematically
I love working with designers who have a product design background but then moved into marketing because they are amazing at this. Since they’re used to designing high volumes of pages with many similar components, they think in a systematic way.
In Figma or other tools, you can create components that can be used in different designs. Ideal case is when you want to update that component (our buttons used to be yellow and now we’re making them green), you can update it in one place and it updates everywhere.
Even if you don’t have it set up to update everywhere, it’s still helpful to think in banks of components.
Different approaches to building your component library
Organize what already exists
The simplest starting point is to take all the design components that already exist and put them in one place (eg Figma). That’s your source of truth.
A centralized component bank alone will make creating anything new so much easier.
Build from motion or video
One cool thing I saw a company do when they were going zero to one and creating a brand for the first time was contracting a motion graphics agency to create a video for them. As part of the video, the agency created a ton of awesome components: boxes, buttons, background textures.
They took all the components from the motion graphics video and turned those into their component bank.
Interactive brand tools
The most flexible version of this is something I’ve seen a lot of design agencies doing more of. Instead of just delivering a static PDF of “brand guidelines,” they turn brand guidelines into a web-based tool where you can check boxes, drag sliders, input specs, and output an on-brand creative asset for something you need. On brand but still unique.
The point is to assemble, not start from zero every time.
When brand lived in my head, every mistake felt personal
Here’s what surprised me most.
When brand lived in my head, every mistake felt personal.
When brand lived in a system, execution felt shared.
Less ego. More engineering.
Brand is taste. But much of taste can be defined. It needs infrastructure.
Otherwise it depends on one person catching everything at the right time. And that’s not scale.
Zero to one is intuition. One to scale is architecture.
I wrote about many of those zero to one lessons here: 9 hard lessons learned going zero to one
That day with the ads was painful. But it forced me to turn brand into something durable. Something repeatable. Something that could move fast without losing trust.