The founder messaged me on a Wednesday. His team was 5 people. By the end of the year, he wanted to hit 15. He’d already figured out which tools did which job: Asana for projects, Drive and Dropbox for files, Frame.io for creative review. But he kept running into the same wall.
Every time someone new joined, onboarding turned into a week of Loom links, Slack screenshots, and “ask Sarah, she knows.” The SOPs that did exist were scattered across Google Docs. There was no single place to learn how the company actually ran. He knew this would break at 10 employees. He wanted Notion to be the brain before it did.
What Was Broken
Their knowledge base was a folder in Drive. Their training system was whoever had time that week. Their SOPs were half-finished Docs from whenever someone got annoyed enough to write one down.
The real problem wasn’t missing information. It was that no one could find the information they already had. A support agent hunting for the refund escalation process would end up messaging the founder. A new hire trying to learn the creative review workflow would get three different explanations from three different people.
This is the point where most founders try to fix it by dumping everything into Notion and hoping structure emerges. It doesn’t. I’ve seen that experiment fail in dozens of workspaces.
The Build
Here’s exactly what I built and why.
I separated SOP databases by department instead of using one shared library.
The founder originally asked whether a single SOP database could serve all departments. My first instinct was no, and I told him so. Shared databases sound efficient on paper, but in practice they become landfills. Customer Support SOPs bleed into Operations, filters get complicated, and people stop trusting the search. Each department now has its own SOP database with role-specific templates. Leadership can still search across all of them from Company HQ, but day-to-day, each team sees only what’s theirs.
I built the training system around completion tracking, not content storage.
A training system that’s just a folder of documents is not a training system. It’s a library. The version I built uses Notion buttons to generate new training programs from templates. Each trainee gets a personal dashboard showing their assigned modules, progress percentages, and pending tasks. Management gets a separate view showing onboarding progress across the whole team at once. When a new hire starts, the founder clicks a button, picks a role, and the right modules get assigned automatically.
I kept Notion out of the production pipeline.
This is where most workspace builds go wrong. People try to turn Notion into everything: project manager, file storage, creative review tool. The founder had already decided Asana owned tasks and Frame.io owned video feedback. So I built the Marketing section to hold strategy, creative concepts, competitor references, and UGC creator databases, with zero production workflow. Concepts get documented in Notion, then handed off to Asana. No overlap. No double-entry. Each tool stays in its lane.
I built verification and ownership into every SOP.
Every SOP page has an owner, a verification status, and a last-updated timestamp. If a policy hasn’t been verified in 90 days, it shows up in the dashboard as needing review. This is the difference between a wiki that stays alive and a wiki that quietly rots while everyone stops trusting it. It’s a small design decision but it’s the one that determines whether the workspace still works in 12 months.
The Result
The company moved into the workspace with a clear map of where every piece of knowledge lives. New hires go through structured training programs with tracked completion instead of ad-hoc explanations and hoping things stick. The founder has one place to check if SOPs are current, who owns what, and where the gaps are.
The workspace is ready to hold 20+ employees without restructuring. That was the entire point: build it right the first time, so growth doesn’t force a rebuild.
On the second review round, the founder asked for one addition: a creative performance database to track hooks, angles, editors, spend, and CPA over time. Added in 12 hours. Everything else landed on the first pass.
What I Learned
The best decision in this project wasn’t technical. It was structural. Saying no to the shared SOP database and pushing for department separation is the kind of call that looks small during the build and pays off for years afterward. Most workspace problems aren’t solved by better features. They’re solved by better boundaries. Notion works best when you decide upfront what job it’s NOT doing.