Theme 3, Article 3: Integrating governance across tiers: coherence as a design requirement


A council is six months into a major service transformation. The programme board sees its risks. The corporate board sees a filtered summary. A neighbouring directorate is planning a structural change that will cut across the transformation's delivery model in three months' time, but no one in the transformation space is aware of it because that conversation has no owner. The risk register shows green. The corporate board has no reason to ask. The collision happens.
That is a governance design failure. Coherent governance is about the quality of what flows upwards through governance tiers and laterally across affected domains; and timeliness is key.
Strip it back and governance is the management of risks by another name: reputational, financial, schedule, missing value realisation targets, the risk of unintended consequences. Every governance decision is, at root, a judgement about which risks the organisation will accept, mitigate, or escalate. So the design of governance is the design of how risk information moves, where decisions get taken, and how the consequences of those decisions are tracked back to outcomes.
For most, the belief is that they have it under control by design. They have an escalation route, a reporting rhythm, a set of boards. Is it just a skeleton? Or do they have the connective tissue that makes the skeleton work as a body.
Vertical integration is harder than it looks. Reporting lines exist but do they carry the right information to the right place at the right time? Information is filtered at every tier, sometimes for good reason and sometimes for reasons passing understanding.
By the time a project-level concern reaches the corporate board, it has been summarised, diluted, or aggregated with other concerns until the original signal is weakened. And siloed risk management at project or programme level means the true level of risk carried by an organisation may not be visible at all. Each delivery unit sees its own risks – the local register. The aggregate picture, the one that leadership really needs to understand, is reconstructed from filtered sources rather than originating sources and, thereby, may not convey the necessary weight of message. Decisions are made on a partial dataset, for want of coherent, integrated governance design.
Underneath this a more foundational problem lurks. If the scoring model does not surface urgency of action, as well as potential severity, the right risks won’t trigger the right escalations, and the right lateral conversations do not happen. A risk that is severe but distant gets close attention. An apparently less severe risk that is more imminent gets less attention or is deferred.
Good governance information architecture starts at source.
Lateral integration is rarely designed at all. Most governance frameworks assume information either moves vertically or it does not move at all. Take the everyday case. One domain tells other adjacent domains what it is planning, so they can decide if there is value in that thinking for them, or something else to be concerned about. That conversation has no owner.
If that conversation happens at all, it happens through established relationships. Where they are good, these connections work informally. Where they are weak, or where events such as reorganisation disrupts them, they break down and go unnoticed until consequences emerge.
Lateral connections sometime surface what vertical aggregation has missed. A risk that one programme has logged, and another has not yet seen, but should have. A design or planning assumption that two programmes are making in opposite directions. An opportunity that one domain is closing off because it does not realise another domain depends on it remaining open. None of this reaches corporate governance through the vertical line because it was not designed to carry it.
Escalation is not delegation. Pushing a problem up the chain without analysis, options, and a recommended position is an omission. The escalating party is invariably closer to the problem than the receiving governance tier; they owe that tier a cogently argued resolution, including positions that may not initially find favour. The first instinct is not always best answer. A board asked to decide between options it has been obliged to formulate in real time, on information that is partial, is being asked to do ground work that should have been done before the point of escalation. Decision quality suffers, as does decision timeliness.
Coherence is a design requirement. Vertical filtering, lateral disconnection, and escalation-as-delegation, can be designed out at the front end. They are difficult to retrofit later, because that requires changing habitual behaviours that have grown around these design gaps. The cost of designing in coherence from the outset is modest. The cost of running without it can be substantial in value lost.
Coherent governance does not happen by accident. It is what an organisation builds when it accepts that designing good governance is designing more effective risk management: how organisations think, decide, and act through uncertainty.
A methodology give you structure. Integrated delivery systems give you safer trajectory and greater momentum.



Comments