Adjacencies & the Plan

Adjacency analysis is the study of how strongly each function in a workplace needs to be near each other function. Some relationships are strong and positive - a design team and the project rooms they live in, reception and the waiting area, the pantry and the social space - while some are strongly negative: a quiet focus zone should be far from the noisy cafe, and confidential HR should not sit in the open thoroughfare. Most relationships are weak or neutral, and recognising that is itself useful, because it tells you what you are free to move when harder constraints push. The point of the analysis is to surface these relationships explicitly, before the plan freezes them in by accident.

The quickest tool is the bubble diagram - a freehand sketch where each function is a circle (a bubble), sized loosely by its area, and lines between bubbles show the relationships: a thick or bold line for a strong need to be adjacent, a thin line for a weak one, and a dashed or marked line for a relationship to avoid. The diagram in this lesson shows the idea: a cluster of bubbles held together and apart by the pull and push of their links. A bubble diagram is deliberately loose and topological - it captures who relates to whom without committing to exact positions - so it is fast to draw, fast to revise, and forgiving. You can sketch five versions in the time it takes to draw one wall in software, and that speed is the point: you are exploring relationships, not drafting.

Where do the relationships come from? Not from the designer's assumptions, but from the organisation itself (the listening of Module 1): who works with whom, which teams hand work back and forth, who needs privacy, where visitors go, what must stay acoustically calm. For anything important, ask rather than guess - teams are full of non-obvious dependencies, and a confident wrong adjacency is worse than an admitted unknown. The bubble diagram then becomes a shared artefact you can put in front of the client and test: 'we have put your engineers next to the labs and away from sales - is that right?' Because it is quick and legible, it invites the conversation that makes the eventual plan correct rather than merely plausible.

A bubble diagram is wonderful for a dozen functions, but organisations have many more, and on a busy sketch the lines start to cross into spaghetti. The systematic companion tool is the adjacency matrix - a grid with every function listed down the side and across the top, so each cell records the desired relationship between one pair. A simple scale works well: a mark for 'must be adjacent', a lighter mark for 'good to be near', a blank for 'no strong need', and a distinct mark for 'keep apart'. Filling the matrix forces you to consider every pair, not just the obvious ones, and it catches the relationships a freehand bubble diagram quietly drops. It also makes priorities explicit, which matters because you can never satisfy every adjacency at once.

That last point is the heart of the craft. In any real floor, adjacencies conflict: two teams both want to be next to engineering, the cafe wants to be central but must be away from the library-quiet zone, everyone wants a window. You cannot grant them all, so adjacency work is really about ranking - deciding which relationships are load-bearing and which can yield. The honest questions are: which of these adjacencies, if broken, actually harms how people work, and which are merely nice? A team that hands work to another every hour has a real adjacency need; two teams that meet monthly do not, however much they would enjoy being neighbours. Protecting the few critical adjacencies and relaxing the many optional ones is what lets a plan resolve at all.

Think, too, about what is being made adjacent - not always whole departments, but often shared resources and the settings people move between. A cluster of focus rooms should be near the teams that need quiet; the printer and supplies near those who use them; the meeting rooms distributed so no one is marooned far from one. And some adjacencies are about flow and sequence rather than teams: a visitor should meet reception, then a waiting area, then a meeting room, in a clean path that never routes them through the working floor. Reading adjacency as the relationships between settings and journeys, not only between org-chart boxes, is what turns a correct diagram into a humane one.

Beneath the specific adjacencies of any project run two organising gradients that apply to almost every workplace, and leaning on them turns a tangle of relationships into a simple spatial logic. The first is loud to quiet. Activities in an office sit on a spectrum of noise and energy: the buzzing cafe, the animated sales floor and the lively collaboration zone at one end; open working desks in the middle; and focus rooms, the library-quiet zone and phone booths at the other. When you arrange these along a gradient - loud at one end of the floor, quiet at the far end, with the transition managed in between - sound behaves itself, because loud and quiet are separated by distance and by the buffer of the moderate middle. Collide them instead, putting a phone booth beside the pantry, and no acoustic treatment will fully save it (the subject of Module 4). The diagram shows this gradient as a band from energetic to calm.

The second gradient is public to private. A workplace runs from the fully public - the entrance, reception, the spaces visitors and candidates see - through semi-public meeting and client areas, into the staff-only working floor, and on to the most private zones: HR, leadership, confidential project rooms, wellness and parent rooms. Organising the plan so that you move along this gradient as you go deeper - public at the threshold, private at the back - gives security, confidentiality and a natural sense of arrival, and it stops visitors from wandering where they should not or sensitive conversations from happening in the open.

The beauty of the two gradients is that they usually align and reinforce each other: the public end of a floor (reception, cafe, collaboration) tends also to be the loud, high-energy end, while the private end (focus, confidential work) tends to be the quiet one. So a single organising move - energy and publicness high near the entrance, calm and privacy deep in the plan - resolves a great many adjacencies at once and gives the floor an intuitive, legible order that people feel even if they never name it. Where the gradients conflict (a private space that is also loud, say, or a public one that must be calm), you handle the exception deliberately with enclosure or buffering. But starting from the gradients means you are arranging a floor along a clear logic rather than placing each function in isolation.

Adjacency diagrams and gradients are not the plan; they are the brief for it. The final craft of this lesson is the translation - turning the web of relationships into a real arrangement on a real floor plate, with its fixed core, its windows, its columns and its single front door. This is where the abstract meets the stubbornly specific, and it is always a negotiation rather than a clean derivation.

The method is to work from the fixed and the critical outward. Start with what you cannot move - the core, the entrance, the structure, and any function tied to a particular place - because these anchor everything. Lay the gradients over the plate next: decide which end is the loud-public end (usually nearest the entrance) and which is the quiet-private end, so the floor has a spine of logic before any team is placed. Then place the strongest adjacencies first - the load-bearing relationships your matrix marked as critical - honouring them even at the cost of the weaker ones, and position shared settings (meeting clusters, focus rooms, the social heart) where the gradient says they belong. Only after that do you fill in the neutral and flexible functions, which by definition can go where there is room. Throughout, you test-fit: sketch an arrangement, walk it in your mind as a user and a visitor, check it against the matrix and the gradients, find where it breaks, and revise. A plan is good when it survives that walkthrough.

Two honest cautions close the lesson. First, you will not satisfy every adjacency, and a plan that claims to has usually either over-simplified the relationships or ignored the building; the mark of skill is knowing which compromises cost little and which would hurt. Second, the plan must still obey the things adjacency cannot override - egress routes, accessible clearances, and the services reality - which are verified with the code and the specialists, not traded away for a nicer bubble diagram. Do this well and the result is a floor that feels coherent and effortless to its users precisely because the hard thinking about relationships happened first, on cheap paper, before anything expensive was built. That is the quiet power of adjacency work: it is where a plan is really designed.