S01E20: What makes every codebase unique? We will build that today.

Let's begin

The last two weeks, we built the Main Branch.

You have five files. They tell your AI who your buyer is, what your product does, how you handle objections, what your process looks like, and how your team sounds. That's the foundation. Every rep on your team could load those same files and start from the same place.

That's the point of the Main Branch. It's shared.

This week, we build the part that isn't.

The Main Branch comes from your company. Your manager could hand it to you. Your enablement team could build it. In theory, it should already exist somewhere in your org.

The Personal Branch is different. Nobody can hand you this. It only exists if you build it yourself. And most reps never do.

What Lives in Your Head

Here's something worth stopping to think about.

Everything you know about your accounts right now lives in your head.

The relationship history with your champion at Acme. The reason you haven't looped in the VP yet. The thing the last rep said on a call that poisoned the well before you took over. The fact that your contact is 90 days into a new role and still figuring out her own priorities.

None of that is in Salesforce. None of that is in the company playbook. And right now, when you open your AI and ask it to help you prep for a call or draft a follow-up, it has no idea any of it exists.

The same is true for your voice. You have a way you write when you're at your best. A way you open an email. A way you handle a follow-up that doesn't sound like everyone else on your team. That's yours. The team voice guide captures how the team sounds. It doesn't capture how you sound.

The Personal Branch is just the act of writing all of this down in a place your AI can read.

The Four Files

Your Personal Branch has four files. Each one captures a different kind of knowledge that only you have.

accounts.md is the running record of what you know about each of your accounts that isn't in any system. The history. The internal dynamics. The champion. The landmine. The thing you said on a call six months ago that you need to remember before you walk into the next one. One entry per account. Updated after every meaningful interaction.

voice.md is the file that makes your AI sound like you. Not like the team. You. Pull three to five emails you've written that you'd send again tomorrow without changing a word. The ones where you re-read them and thought "yeah, that's exactly right." Drop them in. Add three to five phrases you never use. That's the file. The AI learns your voice from your actual writing, not from a description of it.

deals.md is your real-time deal board. Where each active deal stands right now. What changed this week. What the specific next move is. What you're working around. What the close looks like if everything goes right. Update it every Friday. If you skip a week, your AI is working off outdated information. Outdated deal context is worse than no deal context.

territory.md is your read on your market. What you've noticed about your specific segment that the ICP doc doesn't cover. What's shifted in the last 30 days. What keeps coming up across your accounts. The things you'd tell a new rep on their first day that they won't find in onboarding.

Building voice.md

voice.md is the most personal file you'll build and the one most people overthink.

Here is the entire process: open the file, write down three words that describe how you sound when you're writing at your best. Then find three to five emails you've actually sent that nail that. Paste them in. Then write down five phrases you would never use — the ones that make you cringe when you see them in a rep's outreach. Done.

The test is simple. Load the file. Ask your AI to write a cold opener to a prospect. Read what it produces. If you'd change more than three words, your examples aren't specific enough yet. Add more samples. Run the test again.

When the AI produces an opener you'd send without editing, the file is working.

That's the whole thing. Most reps skip this file because it feels abstract. It is not abstract. It's the file that takes everything from "sounds like us" to "sounds like me."

Same Prompt. One More Layer.

You've seen the two-rep comparison before. Let's make it three.

Rep A has no context. Opens a new chat, types a prompt. Gets the opener that starts with "I hope this message finds you well." You know this output. You have written it and deleted it a hundred times.

Rep B has the Main Branch loaded. Buyer profile. Product context. Team voice. Gets something useful. Specific enough to be credible, on-brand, better than most outreach. But it still sounds like it came from the team, not from them. If you swapped the signature, you wouldn't be able to tell who wrote it.

Rep C has both layers loaded. Main Branch plus their accounts.md and voice.md. The output knows who the specific prospect is, references the last conversation, opens the way Rep C opens things. Reads the email and you'd know whose account it was.

Same tool. Same prompt. Completely different outputs.

The personal layer is the diff between Rep B and Rep C. And it's the diff most people leave on the table because they only ever build the foundation.

Keep It Alive

One more thing before you go build.

The Personal Branch is not a one-time project. Unlike the Main Branch, which changes slowly, your personal files change constantly. Your deals move. Your accounts shift. You have new conversations. You write emails that are better than anything in your voice file right now.

• accounts.md: update it after every meaningful interaction. Not a full rewrite. One entry. What happened, what it means, what you need to remember.

• deals.md: every Friday. Where each deal stands. What changed. What's next. Block 15 minutes. Do not skip this.

• territory.md: monthly. What's shifted in your market. What you're seeing across accounts that the playbook doesn't account for.

• voice.md: quarterly, or after a big win. Pull your best recent emails. Add them. Let the file get more specific over time.

A stale personal branch isn't neutral. It's actively misleading. If your deals.md says an opportunity is in late-stage when it stalled two months ago, your AI is going to write like the deal is healthy. That makes the output worse than no context at all.

Build it. Keep it current. That's the whole job.

Rep Action this week

Build voice.md.

Not accounts. Not deals. Voice first.

Pull three emails you'd send again tomorrow. Paste them in. Add five phrases you never use. Load the file into your AI and ask it to write a cold opener for a current prospect. Read what comes back.

That test will show you immediately whether the file is working. If the output sounds like you, you're done. If it sounds like everyone else, add more samples and run it again.

One file. One afternoon. This is the part nobody can give you.

2026 State of AEO Report

A year ago, most marketers weren't thinking about AI search. Now it's one of the fastest moving channels in the industry and nobody has a playbook yet.

So we built one. We surveyed hundreds of marketers to find out how they're approaching answer engine optimization, where they're investing, what's actually working, and what isn't.

The result is the 2026 State of AEO Report. Real data. Real strategies. A clear picture of where AI search is headed and how to get ahead of it.

Download the free report

Moda is the viral AI design agent for polished, on-brand slides, docs, ads, and more. Turn prompts into fully editable designs on a real canvas, then export to PowerPoint, PDF, and more.

Try Moda Free Today