Rachel Woods reduced a 3-week process to 3 hours using AI.

What was the process, and what did she call the document that made it possible?

Answer at bottom of email

Rachel Woods was working at Facebook before most marketers knew what a transformer architecture was.

She built and sold a VC-backed wine e-commerce startup using early GPT technology before ChatGPT was a household word. Now she runs an AI operations agency and a training platform called The Exchange.

She's one of the clearest thinkers Kevin and Norm have ever had on the show about what AI is actually capable of doing in a real business today.

The case study she walks through in the episode is the one that stops the room. An agency she worked with had a 20-page standard operating procedure for vetting influencers. It covered brand safety assessment, cross-platform research across TikTok, YouTube, LinkedIn, and Google, content analysis, and audience quality checks. Thorough, professional. Three full weeks to complete for each influencer.

Rachel's team built an AI system around that SOP. The 20-page document became an AI playbook. Agents were assigned to each research task.

The system ran the same process, the same thoroughness, the same standards, in a matter of hours. Three weeks to three hours. The agency didn't fire anyone.

Rachel's philosophy, stated clearly: "Own the playbook, rent the tech." The SOP, the documented process, is the real moat. Not the AI tool. Not the specific model. The playbook is portable. Platforms come and go. Models improve and get replaced.

But if you've built clear, detailed, repeatable documentation of how your business does its best work, that documentation is the input that any future AI system can use.

The entrepreneurs who understand this now are building a structural advantage that will be nearly impossible to close in two years.

This week pairs Rachel Woods on practical AI operations with Dan Ashburn (Titan Network) on the broader agentic AI world and what it means for performance marketing, community building, and content creation.

The through-line: most businesses are using AI as a writing assistant.

The businesses that will dominate the next decade are using it as an operations system. These are fundamentally different things, and the gap between them is growing.

Watch These Misfits Episodes to Dive Deeper:

Russell Brunson popularized the value ladder as a core funnel architecture.

The Concept: Don't convert a cold prospect to your highest-priced offer immediately.

Instead, ascend them through progressively more valuable (and more expensive) relationships.

First-name personalization is table stakes.

It barely moves the needle anymore.

True email personalization means using behavioral and transactional data to make the content itself feel individually relevant.

Here are practical strategies you should use:

Platforms like Klaviyo let you insert dynamic content blocks that render different product recommendations, copy, or offers based on the recipient's profile. Connect your email platform to your CRM and e-commerce data.

Start with one high-intent behavioral signal, like "viewed product but did not purchase in the last 7 days," and build a triggered, personalized flow around it.

For help with email marketing for your brand, contact dragon.fish

The Brand: Glossier, founded 2014 by Emily Weiss.

The Strategy: Before Glossier sold a product, Emily Weiss ran Into The Gloss, a beauty blog and community. She spent years building an audience and understanding exactly what her future customers wanted from a beauty brand that the industry wasn't giving them. When Glossier launched, it wasn't creating demand. It was fulfilling the demand Emily had already mapped.

The Result: Glossier was valued at $1.8 billion at its 2021 peak. It became the template for community-first DTC brand building, proving that the audience is the most important asset, and the product is its natural fulfillment.

The AI application: Glossier's model maps directly to today's AI-era playbook. Build content. Build community. Mine the community for your product roadmap. Launch to a warm, pre-validated audience. Rachel Woods' approach to building The Exchange followed the same sequence.

The oldest known written reference to whiskey dates back to 1494, buried in Scottish tax records. A monk named Friar John Cor received an order for "eight bolls of malt wherewith to make aqua vitae." A boll was a dry measurement of grain. Eight bolls would have been enough malt to produce roughly 1,500 bottles of whiskey.

That single government record is the first confirmed evidence of distilled spirits in Scotland. The Scotch whisky industry traces its entire heritage to one line in a treasury ledger from 531 years ago. Monks, taxation, and grain.

The industry was born bureaucratic.

The next Collective Minds Society Cigar & Whisky trip is Feb 18-22, 2027

Most marketers use AI reactively: one prompt, one response, repeat. Building a custom agent means your AI understands your business, your voice, and your SOPs before you ever type a prompt. Here is the process:

Sample Prompt (for building the agent's system prompt):

A well-built custom agent does not replace your marketing judgment. It runs the repeatable parts so your judgment can focus on the decisions that actually require it.

Rachel Woods defines an AI operator as someone who translates existing business processes and SOPs into AI playbooks and manages the execution of AI systems. A prompt user types one-off requests.

An AI operator builds systems that run workflows repeatedly and autonomously. The role does not require programming skills; it requires process documentation and systems thinking.

Basic AI automation follows fixed rules: if X happens, do Y. Agentic AI can learn from context, adapt to new information, and execute multi-step workflows autonomously without human intervention at each step.

Dan Ashburn described agentic AI as browser-capable agents that can research, fill forms, analyze data, and complete tasks that previously required a human in the loop.

Rachel Woods found that businesses with documented SOPs got dramatically better results from AI implementation than businesses attempting to use AI without them. The SOP is the instruction set the AI runs.

Without it, AI produces generic outputs. With a detailed 20-page SOP, as in the influencer vetting example, AI can replicate the exact quality standards of your best human work at 10x the speed.

Rachel's "own the playbook, rent the tech" philosophy addresses this directly. The AI tools themselves will be commoditized.

Any competitor can subscribe to the same model. But the documented business processes that define exactly how your company does its best work, those are

THE ANSWER YOU’RE LOOKING FOR

Influencer vetting reduced Rachel’s workload from 3 weeks to 3 hours.

The document that made it possible was the AI playbook, Rachel's term for a 20-page SOP converted into structured agent instructions.

She explained in her Marketing Misfits episode that the existing SOP was the key asset. The AI didn't replace the process; it ran the same process dramatically faster and more efficiently.

Watch the full conversation