Smaller on Purpose: The Executive Case for Breaking Up Your Own Empire
Photo: EXECUTOR, CC0, via Wikimedia Commons
For most of the last century, the answer to the question "how big should a company be?" was self-evident to American executives: bigger. Scale lowered procurement costs, diversified risk, and impressed investors. Conglomerates were not accidents; they were strategy.
That consensus is collapsing. General Electric, the archetypal American conglomerate, spent 2021 through 2024 dismantling itself into three companies. Johnson & Johnson spun off its consumer health business. Kellogg split in two. 3M, Toshiba, Danaher, and a steady procession of others have all concluded the same thing: some parts of their enterprise were worth more apart than together.
This is not fashion. There is a real financial and organizational logic underneath, and executives who dismiss the trend as Wall Street impatience should understand what is actually driving it.
The Arithmetic of Conglomerate Discount
The core problem is that public markets price what they can analyze. A diversified conglomerate trades at a blended multiple that often sits below the sum of its parts — the so-called conglomerate discount. A company whose industrial division deserves a 14x earnings multiple and whose financial services arm merits 9x may trade at 11x blended. Investors cannot easily value what they cannot cleanly see.
Separation fixes the visibility problem. Post-spinoff, each entity gets its own statements, its own management team accountable to its own shareholders, and its own multiple. Academic studies and the record of recent US spinoffs suggest the combined value frequently exceeds the pre-split whole — sometimes substantially.
But the discount is a symptom, not the disease. The deeper question is whether the conglomerate structure was ever earning its keep.
What the Center Actually Contributes
Defenders of scale point to genuine synergies: shared back office, cross-selling, centralized procurement, internal capital allocation. Those benefits are real. The trouble is that they come bundled with real costs that rarely appear in anyone's budget:
Decision drag. Every layer between a business unit and the capital it needs adds time. In markets that move quarterly, a headquarters that decides annually is a tax.
Talent distortion. High performers in a subsidiary of a conglomerate advance by navigating the parent company's politics as much as by winning in their market. Many leave for standalone companies where the scoreboard is clearer.
Capital misallocation. Headquarters inevitably spreads investment across units, and strong businesses subsidize weak ones. Standalone companies must fund themselves or die — a discipline that concentrates resources where returns actually are.
Strategic blur. Customers, recruits, and investors struggle to understand what a conglomerate is. Ask anyone what a focused company does and they can answer. Ask about a sprawling one and the answer is a portfolio page.
When the center's contributions fail to exceed these costs for a given business unit, that unit is carrying dead weight — and its management knows it.
A Decision Framework for the Reluctant Empire-Builder
Most executives will never face a GE-scale decision, but nearly every leader of a multi-division enterprise faces a smaller version: should this business unit be structured, resourced, or eventually separated differently?
Four questions cut to the heart of it.
Would this unit win more as its own company? If its growth is constrained by parent-company priorities rather than market conditions, separation deserves analysis. If it thrives on the parent's brand, balance sheet, or distribution, the case weakens.
Does it share anything essential with the rest of the enterprise? Genuine operational integration — shared technology platforms, overlapping customers, common manufacturing — argues for cohesion. A business that shares only a logo and a pension plan is not integrated; it is merely housed nearby.
Could the market value it more visibly? Run the sum-of-the-parts exercise honestly, with outside advisors if necessary. If the answer consistently comes back 15 to 30 percent apart, that gap compounds into a strategic obligation.
Does leadership have the will to shrink? This is the binding constraint. Divestiture is a confession that the empire will get smaller, and executives are compensated and evaluated on growth. The organizations that execute separations well are led by people who measure success in value created, not revenue consolidated.
Fragmentation as Strategy, Not Retreat
The uncomfortable reframe for today's leadership teams: the org chart is not sacred. A structure assembled through decades of acquisitions, defended through inertia, is not a strategy — it is an inheritance. The conglomerate era built some of America's greatest companies, and the unbundling era may refine them into sharper, faster, more accountable ones.
The executives who thrive in the coming decade will not be those who defend the largest perimeter. They will be those willing to ask, without flinching, which parts of their enterprise would grow better on their own — and to act on the answer before activist investors ask it for them.
