Ask why a building’s junctions are crude and you will get answers about the contractor, the client, the programme, or the standard of the industry. The answer is usually simpler and is written on the appointment: somebody agreed to do the work for a number that does not include drawing the junctions.
This is not a complaint about money. It is a description of a mechanism, and it produces predictable results.
What a percentage actually buys
Take a £14m building and a 4 per cent architectural fee: £560,000 across five or six years, covering concept through completion, split across every stage.
The technical design stage typically gets 25 to 30 per cent of that — call it £160,000. At a blended charge-out rate of £75 an hour, that is roughly 2,100 hours to produce every drawing a contractor will build from. On a building of that size, with a facade, three or four wall types, two stair types, a roof, and every internal junction, 2,100 hours is not generous. It is enough to produce a coordinated set that describes the building competently and not enough to interrogate any junction twice.
At 7 per cent the same stage has 3,700 hours, and the difference is entirely visible in the finished building. It shows up in the places where two materials meet, because those are the drawings that get done last and cut first.
Where the hours go when they are short
The sequence is consistent enough to predict.
- Standard details replace specific ones. A generic parapet detail is used where the actual condition is slightly different, and the difference is resolved on site.
- The repeat conditions get drawn and the one-offs do not. Which is backwards, because a repeated condition will be caught by the contractor’s own setting-out and a one-off will not.
- Coordination becomes reactive. Clashes are resolved when the model flags them rather than designed out in advance, which means the resolution is whichever move is easiest, not best.
- The specification is inherited. A previous project’s spec is edited rather than written, and it carries assumptions about products that are no longer correct.
None of these is negligence. Each is a rational allocation of insufficient hours, made by people who know exactly what they are giving up.
Why fees fell and why they stay down
The proximate cause is competitive tendering of professional services, which arrived in public procurement and spread. When a fee is a scored criterion, the market clears at the lowest number anyone believes they can survive.
The deeper cause is that the value of the work is invisible at the point of purchase. A client can compare two fee proposals precisely and cannot compare, in advance, the quality of two sets of technical drawings. So they compare the thing they can measure. Every profession with this asymmetry has the same problem.
Practices have responded in three ways, and it is worth being clear-eyed about which are sustainable. Some specialise until they are not comparable, which works and is slow to build. Some take the low fee and manage the scope hard, refusing anything not in the appointment — which is honest and produces the buildings described above. And some absorb the difference in unpaid hours, which is the industry’s default and is why so many people leave it.
From a director, 34-person practice “We stopped bidding on fee three years ago. We lost about a third of our pipeline and our profit went up, because the third we lost was the third that was costing us money.”
What a client can do about it
For anyone commissioning a building and hoping for a good one, there are two levers, and neither is complicated.
Buy the technical stage properly, separately if necessary. The concept stage is cheap and is where the excitement is. The technical stage is where the building’s actual quality is determined, and it is the stage most often squeezed because it produces nothing photogenic.
And pay for the site stage. Inspection and site queries are frequently written as a lump sum that assumes a smooth job. On a difficult project the architect either exceeds it unpaid or attends less. Neither outcome is what the client thought they were buying.
Quality in a building is not a matter of talent, mostly. It is a matter of how many hours somebody could afford to spend on the drawing where the roof meets the wall.
