How I Built a Marketing System in One Session

This post was written with AI assistance.

I’ve been running two Twitter accounts — my personal (@jessepeplinski) and my business (@peptechstudios) — and neither had a clear direction. My personal had ~180 followers with no consistent posting. My business had 6 followers and barely any content. Before I started posting on both, I wanted to figure out what goes where so I wasn’t muddying the waters between the two.

So I sat down with Claude Code and built the whole marketing system in a single session. Strategy docs, content calendars, action plans, and a reusable framework I can use for future clients. Here’s how.

The problem

I’m a software engineer building a web dev studio. I have two brands to manage:

Jesse Peplinski — the builder. Projects, blog posts, AI/engineering opinions, the founder journey.
Pep Tech Studios — the business. Client-facing, focused on outcomes for small business owners.

The issue wasn’t that I didn’t have content. My personal site has a projects page, a blog with multiple posts, and interactive demos. My business site has services, pricing, and a full intake flow. The content was all there — I just had no system for turning it into posts, no schedule, and no clear separation between the two brands.

The setup

I created a marketing/ repo and treated myself as my own client. The folder structure is designed to scale — every client gets the same layout:

marketing/
  ACTION-PLAN-TEMPLATE.md
  clients/
    jesse-peplinski/
      strategy.md
      action-plan.md
      content-calendar.md
    pep-tech-studios/
      strategy.md
      action-plan.md
      content-calendar.md

The idea is simple: when I onboard a new client, I copy the template into a new folder and start filling it in. Same structure, same process, every time.

Brand separation

This was the most important decision. I had Claude look at both of my sites to understand the content and positioning, and we landed on a clear split:

Personal = the builder. Projects, prototypes, AI-assisted development, systems thinking, the blog. This is where people follow the journey. People follow people, not brands — so the personal account is the primary growth engine.

Business = the storefront. Services, outcomes, portfolio, tips for small business owners. Plain language, no deep tech. This is where trust gets converted into clients.

The rule is simple: my personal account mentions @peptechstudios naturally when relevant. My business account stays focused on clients. Cross-posts are rare and intentional, not routine.

Strategy docs

Each client gets a strategy.md that covers:

Platforms — which accounts, verified status, follower counts
Brand identity — tagline, mission, positioning
What to post / what NOT to post — this is the most useful part. Knowing what doesn’t belong on an account is just as important as knowing what does.
Tone — my personal is direct and conversational. My business is professional but warm.
Posting cadence — Tuesday and Thursday at 9:15 AM, via Buffer.
Growth levers — specific tactics, not generic advice.

Having this written down means I don’t have to make the same decisions every time I sit down to write a post. The strategy is already decided. I just execute.

Content calendars

I started with individual markdown files per post, but quickly realized a single table per client is way easier to manage:

Date	Status	Pinned	Media	Post
2025-04-14	draft	yes	screenshot	Post content here…

The Status column tracks the lifecycle: draft → scheduled (loaded into Buffer) → posted (published).

The Media column was a key addition. Posts with images or video get significantly more engagement. By planning media alongside the copy, I’m not scrambling for a screenshot five minutes before a post goes live.

I drafted three weeks of content for both accounts in one pass. For my personal account, I pulled directly from existing projects and blog posts — the RAG demo, the MCP visualization, Gridiron Rumble, the blog on modern software engineering. For the business account, I leaned into mission and values — why I started, transparent pricing, the pain of working with agencies — since I don’t have client case studies yet.

Action plans

Each client also gets an action-plan.md — a checklist of everything needed to launch. Website, Twitter, LinkedIn, Facebook, Buffer setup, content. Items are checked off as they’re completed, so I can see at a glance what’s done and what’s left.

There’s an ACTION-PLAN-TEMPLATE.md at the root level that I copy for new clients. Same structure, blank checkboxes, ready to customize.

Platform decisions

We landed on three platforms per brand: Twitter, LinkedIn, and Facebook.

Twitter — both accounts verified. Primary posting channel.
LinkedIn — personal profile for professional/career content, business page for case studies and credibility.
Facebook — personal profile and a business page for local discovery.

Buffer’s free tier supports 3 channels, so my slots go to: Twitter personal, Twitter business, and LinkedIn personal. The rest get posted manually for now.

What I actually did vs. what Claude did

Me:

All brand decisions — positioning, tone, what belongs where
Platform strategy and budget decisions (verified on both, Buffer slots)
Content direction — which projects to highlight, what order, what angle
Copy review and editing on every post

Claude:

Explored both sites to understand existing content and positioning
Built the repo structure and all markdown files
Drafted strategy docs, action plans, and content calendars
Suggested media attachments per post

What I’d tell someone starting from scratch

Treat yourself as your own client. Use the same structure and process you’d use for a paying client. It forces you to be deliberate instead of ad-hoc.
Separate your brands clearly. Decide what goes where and write it down. The “what NOT to post” list is more valuable than the “what to post” list.
Plan media with the copy. Don’t write the post and figure out the image later. A post with a 30-second gameplay clip or a screenshot of your work will outperform text every time.
Start with what you have. I didn’t create new content for my first three weeks of posts. I pulled from existing projects and blog posts. You probably have more content than you think — you just haven’t packaged it for social yet.
Keep it simple. I started with individual files per post, realized it was overkill, and consolidated to a single table per client. Markdown in a git repo. No fancy tools. You can always upgrade later.

The result

In one session, I went from two neglected Twitter accounts with no strategy to a full marketing system — brand separation, strategy docs, three weeks of scheduled content for both accounts, action plans, and a reusable framework for future clients. Both accounts are verified. The content calendars have media planned. The action plans show what’s done and what’s left.

The system lives in a git repo, which means it’s versioned, searchable, and I can hand it to Claude in any future session with full context. When I onboard my first marketing client, I copy the template, fill it in, and the process is already built.

What’s next

There’s more to explore. Instagram, TikTok, and other channels could make sense down the road. The markdown content calendars will eventually move to Google Sheets for easier client-facing views. And the thing I’m most excited about — I want to fully automate the posting pipeline via the Buffer API. Write the post in markdown, and it flows straight to scheduled posts across all platforms without manual copy-pasting. That’s probably a future post.

But for now, this works. Get the system working on a few platforms first, then automate and expand.

That’s the pattern I keep coming back to. The thinking has to be mine. The system-building is where Claude shines.

If you need help with marketing, building a website, or creating systems like this for your business — visit peptechstudios.com.