Organisation Design Basics for Founders

Founders who would never ship an unconsidered product routinely run unconsidered organisations. Teams form around early employees rather than work. Reporting lines record history — who hired whom — rather than logic. Titles inflate to close candidates. By 80 or 100 people, the structure nobody designed is quietly deciding what gets built, what gets ignored, and who burns out.

Organisation design is not boxes and lines. It is the physics of how work flows — and founders who learn its basics make faster, cheaper, calmer companies.

Structure follows strategy — really

The first principle is old and constantly violated: decide what the company must be great at, then shape structure to protect it.

• If speed of product iteration is the bet, organise around small cross-functional squads with real autonomy — and accept duplication as the cost of speed.
• If operational excellence at scale is the bet, functional depth and standardisation win — and accept slower cross-functional coordination as the cost.
• If customer intimacy by segment is the bet, organise around segments — and accept heavier coordination on shared platforms.

Every structure is a trade-off; the design act is choosing your costs deliberately. A reorganisation that doesn't begin with "what are we optimising for?" is furniture rearrangement.

The few rules that prevent most messes

• Organise around work, not people. When a structure is built to give a loyal early employee something to run, the work bends to fit the person. Design the role the work needs; then assess honestly whether the person fits it.
• One person, one boss. Matrix sophistication can wait until you have the management maturity to run it. Dual reporting in a scale-up mostly produces dual confusion.
• Push decisions down, push information up. Structure should put decision rights where the information is. If pricing exceptions, hiring approvals, and small spends all flow to founders, the structure is a bottleneck wearing an org chart.
• Mind your spans and layers. Managers with two reports add a layer without adding leverage; managers with fifteen cannot manage. Most scale-ups need fewer layers and wider, properly supported spans — a topic worth its own attention.
• Design roles before hiring seniors. "Hire a great COO and let them figure it out" imports an expensive person into an undefined job. Define the charter, the decision rights, and the interfaces first.

The founder-specific traps

• Title inflation. The VP title granted at 30 people to close a candidate becomes the ceiling problem at 200 people, when real VPs are needed above them.
• The shadow org. Whatever the chart says, people learn that decisions happen in the founder's WhatsApp. If the structure is to mean anything, founders must route decisions through it — including their own.
• Reorganising around problems. Moving a struggling function under a different leader is sometimes right, but often a way to avoid the performance conversation the situation actually requires.

When to redesign — and how often

Healthy triggers: strategy shifts, persistent cross-team friction, decision bottlenecks at the top, or headcount doubling. Unhealthy triggers: a board comment, a new executive's preference for the structure they had before. Expect meaningful redesign roughly at each doubling — 30, 60, 120, 250 — and treat each as a project with a problem statement, options, and a transition plan, not an overnight chart swap.

Founders do not need to become design experts; they need a working vocabulary and a thinking partner who has seen the patterns. That is exactly the kind of episodic, senior work our fractional HR practice handles — start the conversation before the next doubling, not after the friction.