This is one of the questions I am actively building through L&F CG Commercialization Operating System.
L&F CG is a commercialization company. The common thread is not a specific industry, product or technology. It is the ability to understand a market, find and create demand, shape an offer, build a go-to-market system, sell, deliver, measure and learn.
Sometimes L&F CG does that as an adviser — for example when helping professional-service companies commercialize better. Sometimes it does it as a co-owner — as with Gastropub 2026 and Boutique Hotel 2026.
The question is whether the capability behind those different projects can become cumulative rather than starting over each time.
Current working model
Project → evidence → outcome → learning → method → agent / tool → stronger next project.
The aim is to build L&F CG as a human-led, agent-driven commercialization company. AI is not the identity of the company. It is part of the operating leverage.
I want the parts that require judgment, relationships, trust, creativity, negotiation and consequential decisions to remain human-led. Research, monitoring, data collection, analysis, documentation, coordination, quality assurance and administration should increasingly be supported — and where appropriate performed — by agents and tools.
What we are building
The first layer is a shared commercialization capability model: market intelligence, demand intelligence, strategy and positioning, offer and business model, go-to-market, sales execution, delivery and adoption, measurement and economics, learning and improvement, and portfolio allocation.
Around that, we are building a practical operating system that can answer four questions for any L&F CG project:
What are we trying to commercialize?
What evidence do we have and what is still only a hypothesis?
What is the next gate or decision?
What should we learn here that can make L&F CG better next time?
The first concrete build package is a capability inventory and portfolio map: what already exists, what is duplicated, what belongs to the reusable core, what is specific to a client or venture, and which capability gaps are most worth closing next.
Two applications of the same capability
Client commercialization: L&F CG is paid to improve another company's commercial system.
Venture commercialization: L&F CG participates as an owner or development partner where its commercialization capability can materially increase the value of the venture.
The economics and ownership are different. The underlying commercialization logic should increasingly be shared.
An important boundary
Not everything I work on personally belongs inside L&F CG. Private projects such as Jaktradioen remain separate unless there is an explicit decision to make them an L&F CG engagement.
That boundary matters because the aim is not to make L&F CG a container for everything I find interesting. It is to make it exceptionally good at commercialization.
What I currently think
The long-term advantage may come from combining human judgment with reusable methods, governed data, agents and tools — then allowing real project outcomes to improve that system continuously.
That is different from building a large collection of software products. A tool earns its place when it makes L&F CG better at commercialization. An agent earns its place when it can take useful work off the critical path without weakening context, quality or control.
The company should therefore become more capable without its administrative burden or headcount having to grow proportionally with the number of projects.
What I still do not know
Which capabilities should be centralized and reusable across every project, and which should remain domain-specific? Where does human judgment continue to create the most value? Which repeated tasks are mature enough to become agents or tools? How much learning can safely generalize between clients and ventures without losing context or crossing confidentiality boundaries? And which capability investments actually improve commercial outcomes rather than merely making the system look more sophisticated?
Those are the questions this project is now designed to answer.