The AI-Human collaboration behind our content (Part 2)
Last week, we showed you how we generate ideas at the firm and communicate them to founders & our LPs. We highlighted our process: five distinct meeting types and specialised AI workflows that utilise quarterly narratives and dedicated Claude projects tailored to each audience.
If you haven’t read Part 1, you can read it here.
We covered the entire setup: how we capture insights from portfolio reviews and founder conversations, organise them within our narratives, and create AI collaborators loaded with our complete content history, audience preferences, and investment thesis. We're also voracious readers, and by using Reader, all the team members constantly highlight insights, counterarguments, and connections that challenge our existing thinking.
Today, we'll walk you through what happens next: how those captured ideas actually transform into the newsletters, founder resources, and social posts you see from Good Capital.
Mapping ideas to channels
After our various meetings generate dozens of potential ideas each week, our marketing team faces a coordination challenge: which ideas deserve development, and for which audiences?
We use Granola AI to capture all our meetings. Marketing then reviews these captured ideas to identify which ones have content potential. The richest ideas also come from synthesis - connecting a research paper one team member highlighted about consumer behavior with a portfolio company's unexpected pivot that another member flagged. Our meetings become laboratories for testing these connections, and hence Granola becomes invaluable for preserving these intellectual threads.
Marketing then reviews these captured ideas and debates to identify which ones have genuine content potential. Each promising idea gets evaluated for audience fit:
LP Audience looks for strategic insights, investment implications, and market analysis that help understand our thesis and India's economic evolution
Founder Audience needs tactical insights, implementation lessons, and operational guidance they can apply to their own businesses
Social Media Audience responds to quick insights, bite-sized content, and broadly engaging observations about Indian markets.
Importantly, the same core observation often works across multiple channels, but needs to be packaged differently for each audience. A portfolio company's AI implementation could become a pattern-uncovering anecdote for LPs, a tactical guide for founders, or a quick insight for social media.
Once we determine audience fit, we take the specific Granola transcript, meeting notes, and Reader highlights containing that idea and feed them to the appropriate Claude project. Instead of treating this as a bulk data transfer, it's targeted context delivery that gives each AI project exactly the background it needs.
The content creation process
Stage 1: Initial Analysis - Claude starts by reviewing our Granola transcript alongside everything it knows about our previous content and audience preferences. It's particularly good at spotting connections - "This is similar to your manufacturing thesis from ITR 73" or "This could build on the AI intermediary argument from last month." We use these insights to understand how the new idea fits into our broader narrative.
Stage 2: Theme development: This is where we do the creative heavy lifting. We develop and define the core idea, while Claude suggests different ways to frame our argument and points out which approaches have worked well with our audiences before.
Stage 3: Drafting: Once we've nailed the theme, the actual writing begins. We provide the insights and direction while Claude handles the organisation and language flow. When we need supporting data, Claude also helps in the research through a Perplexity MCP.
Stage 4: Iterative refinement: The final stage focuses on packaging - finalising content, considering visual support like charts or infographics, and ensuring the piece is ready for publication. Claude helps us with these creative direction decisions to make the content as engaging and accessible as possible.
Why this workflow works
This content is more consistent, strategic, and scalable than what we could produce through traditional methods - while unlocking more time for our team to research, learn and creatively wrangle anecdotes into powerful ideas. Every newsletter, social post, and founder resource emerges from this collaboration, proving daily that the most powerful AI applications amplify human capabilities rather than substitute for them.
More importantly, it demonstrates the same principles our portfolio companies use to transform their industries: AI as an intelligent coordinator that preserves and enhances human value creation rather than replacing it.