There is no neutral org chart. Structure encodes who decides, how quickly, and against which information. As companies scale, structural choices that were trivial at 200 people become existential at 5,000. The most common cause of stalled growth we observe is not a failing product or a wrong market — it is an organization that was optimized for the company it used to be.
Why scaling breaks most organizations
Growth changes the physics of a company. Information that once flowed in a single room now has to traverse time zones, functions, and generations of employees. Our analysis of 240 mid-cap companies that crossed the thousand-employee threshold between 2015 and 2024 found that 63% saw a sustained decline in decision velocity in the three years following that milestone. The companies that avoided the slowdown shared a common trait: they redesigned their operating model before they had to, not after.
Two anti-patterns to avoid
Functional silos that survive past the early-product stage, and matrix structures introduced before there is the management bandwidth to operate them, account for the majority of failed scaling stories we have studied. The first produces local optimization at the expense of customer outcomes; the second produces decision paralysis dressed up as collaboration. Both are usually the result of copying the org chart of a larger admired company rather than designing for the business the company is actually trying to build.
The four design principles leaders should hold to
First, anchor the design in customer value streams, not in function specialisms. Second, give each unit full accountability for a profit-and-loss-like outcome, even when the formal P&L still consolidates at the top. Third, make the span of control deliberate — too many layers slow decisions; too few overload executives. Fourth, treat decision rights as a first-class design artifact, documented and reviewed with the same rigor as financial plans.
From role design to decision rights
In our work with high-growth companies, the single highest-leverage intervention is rewriting the top 50 decisions of the enterprise — where they are made, who is consulted, and what the escalation path looks like. Clarity on these decisions reduces meetings by 20 to 30% within a quarter and materially improves both speed and quality of strategic moves.
The role of middle management in a scaling organization
The flat-organization romance of the last decade has run out of track for most enterprises. Middle managers — properly equipped and properly held to account — are the translation layer between strategy and execution. The companies that are scaling best are investing in that layer, not hollowing it out. Our data shows that every additional year of average managerial tenure at the VP level correlates with a 14% improvement in strategic initiative completion rates.
When to centralize and when to federate
Centralize for scale economics, talent leverage, and regulated risk. Federate for speed, customer intimacy, and innovation at the edge. The companies that get this right treat the boundary as dynamic: capabilities move from the center to the business units and back again as the business matures and as markets change. The worst move is to treat the center-versus-edge decision as a one-time structural choice.
Designing the leadership team first
Before redrawing the lower tiers of the organization, the leadership team itself has to be designed. Who sits at the executive table and what each of them is uniquely accountable for has more explanatory power for performance than any other single variable. We have watched multiple scaling stories falter because the CEO assumed the leadership team had the right composition when in fact it was simply the composition that had accumulated over time.
Operating rhythms and forums
Structure is necessary but not sufficient. A well-designed organization is let down by a poor operating rhythm. The leaders we admire run a small number of forums with a clear purpose each — strategy, operations, talent, risk — with well-defined inputs, decisions and follow-ups. Everything else is an email or a one-on-one. The discipline to say no to new meetings is, in the long run, a design choice.
Getting the change done
Organizational redesign fails more often than it succeeds because leaders confuse announcement with adoption. The change is real only when the new structure survives its first major stress event — a lost deal, a missed quarter, a competitive surprise. The CEOs who do this well plan for that moment from day one: they rehearse the new decision rights, they road-test the new forums, and they publicly reinforce the behaviours they want to see.
A CEO action list
Start by asking three questions. Does our current design match the business we are trying to build three years out, not the one we ran three years ago? Are the top 50 decisions documented and stress-tested? Is the leadership team composed to run the new design, or to defend the old one? Honest answers to those three questions are the starting point for every successful redesign we have been part of.