Strategy Summit 2026: Inventive Strategy and the ‘Unbossed’ Organization

Why It Matters

Identifying a strategic center and flattening hierarchies enables companies to allocate capital faster and capture value in the rapidly dematerializing economy, a necessity for staying competitive in AI‑enabled markets.

Key Takeaways

• Economy shifting from physical assets to intangible digital services
• Companies must identify a strategic "center" for resource allocation
• Novartis' focus on inventive medicines grew market cap to $510B
• Unbossed structures empower small, autonomous teams for rapid innovation
• Treat AI like electricity: give employees tools to experiment

Pulse Analysis

The past half‑century has seen a relentless migration from tangible assets—factories, machinery, and raw materials—to intangible drivers such as patents, software platforms, and data analytics. This dematerialization reshapes the rules of competition: barriers to entry are no longer walls of steel but ecosystems of code and network effects. Analysts now observe that traditional strategy frameworks, which emphasized long‑run cost advantages and physical scale, are losing relevance. Companies that fail to recognize the shift risk allocating capital to legacy assets that no longer generate sustainable returns.

McGrath’s answer is to locate a firm’s “center”—the core purpose that guides investment and talent decisions. Novartis exemplifies this approach: under CEO Vasant Narasimhan, the company shed eye‑care, consumer products and generic‑pharma units, trimming headcount from roughly 104,000 to 76,000. By concentrating on inventive medicines, its market capitalization leapt from an estimated $125 billion to about $510 billion, a more than four‑fold increase. The “centered” model simplifies portfolio choices, allowing resources to flow toward high‑impact, science‑driven initiatives such as radioligand therapies.

The operational lever that makes a centered strategy feasible is an “unbossed” organization—flat, loosely‑coupled teams empowered to experiment without excessive approvals. Small groups of three to twelve members keep coordination costs low while preserving diversity of thought, a formula that Netflix’s Chaos Monkey and Adobe’s Kickbox programs have popularized. When applied to AI, the same principle suggests treating machine‑learning tools as utility infrastructure, giving every employee access to models and data pipelines. Firms that adopt this structure can accelerate innovation cycles, reduce time‑to‑market, and better align autonomous effort with the overarching mission.