Draft status: This post is outlined and ready to be completed. The structure, arguments, and CTA are set. Fill in with specific stories, client examples, or case data before publishing.
The math nobody does
3 hours/week searching for research you’ve already done.
That number comes from asking 50 consultants how much time they spend each week looking for things , not creating new things, just finding things they know exist. The honest answers cluster around 2-3 hours.
At $200/hour billing rate (conservative):
- 3 hours/week × 50 weeks = 150 hours/year
- 150 hours × $200 = $30,000 in annual capacity
[EXPAND: Tell a specific story here. A consultant who, before a proposal call, spent 90 minutes searching for research they knew they’d done on adjacent topics in a past engagement. The files were in four different places. Three had confusing names. One was in Google Docs, one in Notion, one in Obsidian, one emailed to themselves. Eventually found 60% of what they needed.]
Why your notes app isn’t the problem
The consultants most likely to build a second brain are consultants who have already tried Notion, Obsidian, Roam, or Evernote. And abandoned them.
This isn’t a tool problem. The apps are fine. The problem is:
- Maintenance requires discipline you don’t have at 10pm after a client call. You save the article with the intention of tagging it properly later. Later never comes.
- Tags don’t synthesize. Even a perfectly tagged notes app still requires you to open every tagged note and read it. 47 notes tagged “pricing” is still 47 notes to synthesize.
- Different tools started winning different years. Your 2021 notes are in Notion. Your 2023 notes are in Obsidian. Your 2025 notes are in… Claude’s Projects? Now you have four systems.
[EXPAND: Your experience here, or client experience. The fragmented tool graveyard.]
The 15-minute setup
Getting started with the LLM Knowledge Base Pattern is, genuinely, 15 minutes of setup:
- Create a folder on your computer (or Google Drive). Call it
knowledge. - Inside it, create three subfolders:
raw,wiki,outputs. - Start dropping things in
raw. Everything. Google Docs exports, Obsidian vault contents, downloaded PDFs, Notion exports. Don’t organize.
That’s the 15 minutes. The rest is AI’s job.
[EXPAND: Be specific about what “the rest” means , either walking through how to run a compilation yourself with Claude, or explaining how we do it for you. This is where the service CTA lives naturally.]
What you get on the other side
After the first compilation:
- Every topic you’ve ever researched has a concept article, synthesizing what you actually know about it.
- Every source you’ve saved has a one-page summary. Key points. Author’s argument. What’s actually useful about it.
- A master index so you can navigate by topic, not by file name.
- An answer to “what do I know about X?” in under 30 seconds, not 90 minutes.
Before your next proposal call, you query: “what’s my research on [client’s industry] + [problem type]?” You get a sourced research brief. Not 47 links. A brief.
[EXPAND: Ideally a specific Q&A example here. What a query looks like, what the output looks like. Make it concrete.]
The setup case vs. the retainer case
The setup is a one-time investment: $1,500, 48-hour turnaround. You get the first compilation done right.
The retainer is where the ROI compounds: $300/month. New reading compiled every month. The system knows more and more of your domain as time passes.
By month 12, you have a knowledge base that represents a decade of expertise, queryable in real time.
[END with CTA to /second-brain/setup or /for/consultants]
Want this done for you in 48 hours? Book a setup call.