You can't adopt AI without redesigning work.

Understand AI capabilities.
Redesign the work.

AI Adoption Journey — from using AI to do your pre-AI job slightly faster to redesigning end-to-end flows around AI capabilities

Two kinds of AI adoption. One makes individuals faster. The other makes organizations different.

Individuals can handle the first. Only leaders can handle the second. They're the only role that carries both — what AI can do, and what the organization has to become.

Your pre-AI business ran on two forces.

Pre-AI two-force frame — Business Process and Human Capital as the strong axis, AI Capability disconnected at the top

Every business before AI was built on two forces: structured process and the humans who execute it. Org charts, training programs, operating cadences — all optimized for that pair. That was the playbook.

AI is a new force. It can't be trained like people. It can't be documented like process. It has its own economics.

You don't adopt it alongside the old two. You redesign around all three.

Lots of magic, but no single magic bullet.

Three-force triangle — AI Orchestration along process-capability axis, AI Chat along humans-capability axis, AI Features along process-humans axis, AI Gap in center

Most AI tools are strong on one axis. Orchestration (Clay, n8n) bridges process and capability. Chat (ChatGPT, Claude) bridges humans and capability. CRM and sales features (Gong, Salesforce, HubSpot) bridge process and humans. Real power on every axis. None alone closes the triangle.

Knowing which bridges close which gaps in your business is the work. It's a leadership question, and no vendor will answer it for you.

Shared context systems live at the center.

Context is the slice of your strategy your AI needs to be useful — product, buyers, positioning, proof points, brand voice. Most teams patch it themselves: prompt docs, shared Claude projects, point-solution context stores. Each patch compounds. Every strategic change breaks something downstream. Context debt.

Context is debt until it becomes a system. Each domain has an owner, a structure, and a cadence that matches how fast it changes. GTM bi-weekly. Brand quarterly. Campaigns monthly. Events weekly. Customer continuously. One update propagates everywhere.

Shared context sits at the center of the triangle. Leaders build it.

Redesign work around the customer.

Create → Equip → Deploy and Analyze → Customer — the four cycles of redesigned GTM work, with shared context and systems inside Create

Redesigned work shows up as four cycles.

Create. Build and maintain the shared foundation — context systems, applied capabilities, the catalog of skills and orchestration everyone else runs on. New roles live here: GTM Strategy, Product Marketing, Applied AI, GTM Engineering.

Equip. The frontline stops receiving tools. They get context-aware skills they modify on the fly in the customer moment. Frontline AI.

Deploy & Analyze. Customer-facing work generates signal. Signal feeds back into context. Every loop fuels the next.

Individual AI lift is linear. Organizational loops compound. Procurement gets you the first. Redesign gets you the second.

Sam Gong

Currently SVP Marketing at WorkSpan, where I've spent the last 2.5 years building the systems this project is about — a semantic knowledge graph for GTM context, patterns for using Claude, code, co-work, and chat across different roles, AI-powered enablement, AI-powered account engagement, and AI orchestration. The patterns exist because they had to.

Fresh Context is what happens when you take those patterns out of one company and try to generalize them. A writing project and a point of view. The work is in the essays.

Writing in a personal capacity.