Published March 2026
'Mental Scaffolding' is my term for the conceptual architecture that determines what a decision-maker sees as a trade-off, what they consider relevant, and which options feel natural before any specific policy question lands on their desk.
In short: the decision-maker doesn't think "I learned this from so-and-so and here's the trade-off." Instead, they think "this is obviously how the issue works," and ideally feel that they came up with the solutions to fix it.
This is borne out of my view that often, the most effective way to influence a leader is to target them not at the point of decision, but to go further upstream — to change how they think months or years before. Inserting information at the top of the stream takes different forms, each bespoke to the context. But I would say a couple of factors are usually at play: ego, incentives, empathy for why they already think how they do, and an effort to step into their mental architecture.
Sometimes a leader shifts position when an issue is reframed as a legacy-defining project, one where they'd be ahead of the curve, pitched with enough granularity to feel actionable rather than aspirational. For example, the best politicians are historically literate and are therefore acutely sensitive to how they'll be remembered, and most policy reaches them framed in terms of immediate political risk rather than long-horizon positioning. (I think it's interesting and underscrutinised how often President Trump has mentioned in passing not getting into heaven during this Presidency.)
Sometimes influence operates through small-group consensus. A handful of people in a room — sometimes from different organisations or companies — discuss an issue well before it enters the formal policy pipeline or company strategy document, and a position is essentially settled there and then, again shaped by social pressure operating on people who sometimes consider themselves immune to it. These moments are hard to quantify precisely because they are designed to be undocumentable. I see this in AI frequently. It has also happened in scientific research before too.
But you can also influence upstream by building a dam downstream: identifying a chokepoint, applying populist pressure, and forcing upstream attention as their architecture space becomes flooded. This is a well trodden path from the Gracchi brothers to Aristophanes' plays. As pressure builds at the dam, those further upstream — ministers, advisers, CEOs, and so on — must redirect their focus toward the issue whether they intended to or not to clear the flooding. The mechanism is cost, in making inaction more expensive than action.
I think, then, we can break it into four elements:
Scaffolding is epistemic — it changes what someone knows. It demands intellectual patience and the willingness to write for fifty key decision-makers rather than fifty thousand followers, and for your work to never receive public credit.
Legacy capture is motivational — it changes what someone wants. This requires understanding a leader's self-image well enough to repackage policy or ideas in terms they find personally compelling on an almost subconscious level.
Room consensus is social — it changes what someone feels permitted to think. It demands high quality social access — often built on reputation — and the credibility to be heard among peers, and critically a sense of discretion.
Dam building is coercive — it changes what someone must respond to. It requires populist instincts and the stamina to sustain public pressure across months or years and take flak from upstream elites and institutions.
All this to say that in some instances, the most powerful influence is the kind nobody remembers receiving.