On how decisions compound, and how early framing shapes what becomes possible.

On maturity and uncertainty

Maturity is not a property of organisations. It's a property of decisions.

A twenty-year-old funding body launching a new programme faces many of the same open questions as a three-person startup: what does good look like, what tests actually matter, how much to commit before the shape of the problem is fully understood. Meanwhile, young ventures can arrive at premature maturity — constrained early by funding design, governance choices, or incentive structures that were set before the thesis was tested.

Most of the interesting decisions sit in this middle register, where neither standard advisory nor standard investment logic applies cleanly. Advisory work assumes a stable problem and an optimisation context. Investment work assumes a clearly bounded opportunity and a thesis already written. Both leave unattended the question of whether the problem has been framed well in the first place.

We pay attention to those boundary conditions. In practice this means treating maturity as something that applies to specific choices, not to the organisation around them. A mature organisation can make young decisions. A young one can make mature decisions. What matters is whether the decision is taken with an appropriate relationship to the uncertainty it actually faces — rather than the uncertainty the operating model would like to pretend it faces.

A common failure mode is the confident application of mature frameworks to premature contexts. Scaling templates, KPI dashboards, and standard governance packages can obscure the fact that the underlying thesis has not been tested. The organisation produces motion without producing learning. The opposite failure is treating every decision as early-stage when some are not, and endlessly re-opening settled questions under the cover of exploration.

The practical discipline is separating what is genuinely uncertain from what is merely uncomfortable. Uncertainty is structural. Discomfort is contingent on the actors in the room. Decisions should be sized to the first and not the second.

On capital and commitment

No instrument of early capital is neutral.

Whether public or private, early money encodes expectations about speed, risk, ownership, and control. A venture term sheet commits the recipient to a growth curve. A grant commits them to a reporting cadence. A blended instrument commits them to explaining themselves in two different vocabularies. None of these are wrong. None are free. Each one narrows the option space before any substantive work begins.

The question worth asking early is not where to raise capital, but what the capital will oblige the organisation to become. Instrument design runs ahead of strategy more often than is acknowledged. By the time the first tranche lands, a set of choices has already been made on the recipient's behalf — about horizon, about permissible risk, about who counts as a counterparty.

This matters most in environments where standard venture heuristics do not fully apply. Public-interest initiatives, public-private mandates, early research translation, and infrastructure-shaped ventures all encounter capital that does not map cleanly onto a growth-or-die frame. In these environments, choosing the instrument is closer to choosing the governance — and reversing it later is rarely cheap.

Our work on this is forward-looking. Given the instrument being proposed, what will the organisation be five years from now, and is that a destination anyone has actually agreed to? The instrument is usually the strategy, whether or not it has been named as such.

Restraint in capital timing is underrated. Taking the right amount at the wrong time is a common way to lose the next five years.

On structure before scale

Growth is easier to ask about than structure.

The question of what an initiative should become is easier than the question of what about it should stay constant. Growth exposes whichever part of a structure was held together by informal coordination, founder discretion, or the implicit culture of a small team. When those supports reach their limit, the organisation either rebuilds quickly or continues at scale with weaknesses it no longer has the attention to notice.

We prefer to ask, early, what should remain invariant as an initiative grows. Four things are usually worth naming: decision rights, meaning who can commit the organisation to what, and at what threshold. Incentive alignment, meaning whether the rewards still point at the mission when volume increases. Accountability, meaning who answers for what, to whom, on what cadence. And the capacity to adapt without losing coherence.

The failure modes are familiar. Structures designed for ten people survive to fifty through willpower and then fracture. Governance designed for one product does not translate to two. Roles defined around early-stage ambiguity harden into territorial lines. None of these are catastrophes in isolation. Compounded across a few years, they decide the shape of the organisation more than any explicit strategic choice.

Sometimes this analysis leads to growth. Sometimes to deliberate restraint — a decision to stay smaller for longer in order to stay coherent. Both can be correct. The mistake is treating scale as the default and restraint as a confession.

On discretion

Some work is stronger for being quiet.

Some engagements have to be public — coalition building, regulatory influence, anything that needs a public mandate to function. Others are effective precisely because they aren't. Internal strategy reviews, governance redesigns, quiet repositioning before a public move: these are often damaged by premature disclosure. The decision at stake becomes harder to make well once it is being performed.

Professional services tend to equate visibility with effort. The deliverable has to be thick. The presentation has to be long. The activity has to be legible to someone who was not in the room. This is a tax paid in clarity. Decisions that are clean internally become muddled by the performance of them.

We work in both modes and don't assume that exposure correlates with impact. Some of the engagements we value most leave no external artefact. The test we apply is whether visibility would change the quality of the decision — not whether it would change the perception of the engagement.

Discretion is not secrecy. It is a case-by-case call about which audiences add signal and which add noise, and a willingness to make that call rather than defer to whichever habit the organisation has fallen into.