The common problem
Every multi-site business eventually confronts the same underlying governance question: which decisions belong to the centre, which belong to the site, and what happens when the two disagree. Single-site businesses can leave this question comfortably unresolved for a long time, because there is only one site and the question rarely comes up in a way that demands an answer. Adding even a second location forces the question into the open — and how it presents changes considerably as the number and spread of sites grows.
How it presents with a small number of sites
With two or three sites, the governance gap tends to be personal rather than structural. Decisions that used to be made by walking across the room are now made over a call, or not made at the same speed at all, because the person with authority is physically at one site and the question originates at another. The organisation typically has not yet built any formal decision-rights structure at this scale, because the informal approach that worked with one site still mostly works — just with more friction and more delay than before. The risk here is less about outright confusion and more about a quiet accumulation of small inefficiencies that nobody treats as urgent enough to formalise.
How it presents at regional multi-site scale
Somewhere between roughly five and twenty sites, research on decision rights suggests the problem changes character. Bain & Company’s research, led by Paul Rogers and Marcia Blenko, identified a recurring bottleneck specifically between “centre and business unit” — a description that maps closely onto the tension a regional multi-site business experiences between head office and individual locations. At this scale, informal resolution genuinely stops working, because there are now too many sites for the centre to personally track and too many site-level decisions competing for the same limited central attention. Sites that received quick, informal answers when there were only two or three of them now compete in a queue, and different sites often receive inconsistent answers to functionally similar questions, simply because different people at the centre are fielding them without any shared reference point.
How it presents across borders
For businesses operating across national or regulatory borders, the same underlying tension takes on an additional layer, because the “global versus local” bottleneck identified in the same Bain research becomes active alongside the centre-versus-business-unit tension already present at regional scale. A decision that is straightforward to centralise within a single regulatory environment may be genuinely inappropriate to centralise across borders, because local legal, cultural or market conditions mean the same decision cannot sensibly be made the same way everywhere. Cross-border multi-site businesses therefore face not one governance question but two overlapping ones — how much to centralise, and which decisions should never be centralised regardless of scale — and conflating the two tends to produce governance structures that are either too rigid for local variation or too inconsistent for coordinated oversight.
What explains the variation
The consistent thread across all three contexts is that informal, relationship-based decision-making has a genuine capacity limit, and each stage of multi-site growth exceeds that limit in a different way. A small number of sites strains it through simple distance and delay. Regional scale strains it through sheer volume of competing decisions. Cross-border scale strains it by adding a dimension — legal and cultural variation — that centralised judgement alone cannot resolve, however well-resourced the centre is.
A common misconception, visible across all three contexts, is that the appropriate response to strain is always more central involvement — tighter oversight to compensate for growing distance. In practice, the research on decision rights suggests the opposite response is often needed: as capacity strain increases, the specific decisions retained centrally should typically narrow, not expand, so that central attention concentrates on the decisions that genuinely benefit from it rather than being spread thin across every decision a growing site network now generates. A centre attempting to retain the same breadth of decision involvement across three sites, twelve sites, and forty sites across three countries is applying an approach calibrated for the first scale to problems that have long since outgrown it.
It is also worth noting that consistency and centralisation are not the same goal, though multi-site businesses often conflate them. A decision can be made consistently across sites without being made centrally — through a shared, well-understood principle applied locally — just as a centrally made decision can still be applied inconsistently if local conditions are not genuinely accounted for in how it is communicated. The goal that actually matters as sites multiply is usually consistency of outcome, not consistency of decision-making location, and the two require different structures to achieve.
Looking for your own sector? The Other Sectors page points to the closest match.
Questions worth considering
- Which of these capacity limits is the organisation actually approaching — distance and delay, competing volume, or cross-jurisdictional variation — since each calls for a different kind of response?
- Are decisions currently centralised for consistency’s sake being centralised because central judgement is genuinely better suited to them, or simply because nobody has revisited the assumption since the organisation added its most recent sites?
More on this topic: Governance.
Your team is facing similar challenges?
Subscribe to The Riddle Ledger
Perspectives are sent as they are published, with no promotional email, and you can unsubscribe at any time.