Most extension projects go wrong in the first fortnight, before anyone has drawn anything. The decisions made then set the budget, the planning route and how the finished space feels, and good house extensions London is mostly about getting them in the right order.
Here are the eight that matter, roughly in the sequence they should happen.
1. Establish What the Plot Allows Before Deciding What You Want
This is the reversal most people get wrong. They decide on the extension, then check whether it’s possible.
Do it the other way round. Confirm conservation area status on your exact address. Check for an Article 4 direction separately, because these are different controls that often overlap. Pull your planning history through the council portal, since previous permissions sometimes carry conditions removing permitted development rights.
Then check what previous owners built. Every permitted development measurement is taken from the original dwelling as it stood in 1948, or as first built if later. A previous extension counts against your allowance.
An hour of this changes what’s worth designing. Skipping it means designing twice.
2. Decide Whether You’re Buying Width or Depth
Extensions solve one of two problems and people rarely name which one they have.
Depth comes from a rear extension. Width comes from a side return infill or a side extension. A wraparound buys both and costs accordingly.
The test is what’s actually wrong with the ground floor now. If the answer is that the kitchen is too narrow to move around in, depth won’t fix it. If the answer is that there’s nowhere for a dining table, width won’t either.
Get this wrong and you build a bigger version of the room you already disliked.
3. Set Depth From the Plot, Not From the Regulations
Under GPDO 2015 Class A, a single storey rear extension is capped at 3 metres for an attached house and 4 metres for a detached one, where permitted development rights survive. The Larger Home Extension Scheme allows 6 and 8 metres through prior approval, subject to neighbour consultation.
Those upper figures get treated as targets. They’re limits, not recommendations.
Two things go wrong at maximum depth. The middle of the plan goes dark, because glazing at one end can only reach so far. And the garden gets short enough that buyers notice.
Officer reports use the phrase overdevelopment of the plot, and what it usually means in practice is too little garden left, or an extension that reads as dominating the original house.
4. Work Out Where Daylight Is Coming From
Rear glazing lights the last few metres of a room. Beyond about 4 metres of depth, it stops doing the job on its own.
Overhead glazing lights the middle. A roof lantern, or a run of rooflights over the deepest part of the plan, is what makes a deep extension feel like a room rather than a corridor.
This decision has to happen before the structure is set, because rooflight positions depend on where the beams land. Deciding it afterwards means either compromising the glazing or reworking the structure.
5. Settle the Structural Strategy Early
Removing the whole rear wall gives an uninterrupted space and a large steel beam carrying the floors above. Beam cost rises sharply with span.
There’s a design argument alongside the cost one. A long open plan room with nothing breaking it up reads as a corridor with a kitchen in it. Retaining a section of original wall as a pier reduces the span, gives you somewhere for tall storage, and divides the space visually without closing it off.
On period houses the structural survey matters more than people expect. Victorian terraces often sit on foundations 400 to 600 millimetres deep, and party walls in half brick are common. Both change the approach, and both are far cheaper to find in week one than in month five.
6. Fix the Threshold and Floor Level
A flush or near flush threshold between the internal floor and the garden does more for the sense of space than an extra half metre of depth. When the eye reads the floor as continuing outside, the room borrows the garden.
It needs designing early because it affects the slab, the drainage channel at the door and the damp proofing detail. Retrofitting it is expensive and sometimes not possible.
Where the existing soil stack sits matters here too, since moving drainage can change the floor build up and occasionally the level itself.
7. Decide the Layout Last, Not First
This is the decision most people start with, and it should come near the end.
Kitchen layouts are the easiest thing to change and the last thing that should be fixed. A layout chosen in week one gets redrawn when the beam sizing comes back, or when the rooflight positions move, or when planning caps the depth.
Once the shell, the structure and the daylight strategy are settled, the layout follows almost automatically. Working the other way produces the dark long room nobody wanted.
The one layout principle worth setting early is where the kitchen sits relative to the light. Running it along a long wall, with the prep zone under overhead glazing, means you cook in daylight and the storage takes the darker end. A good kitchen extensions ideas will cover the fit out detail, but that positioning decision belongs to the shell design.
8. Choose the Planning Route Deliberately
Two routes exist and neither is automatically better.
Permitted development is faster where it applies. No design assessment, no officer opinion, and a lawful development certificate available afterwards to confirm the position for future buyers.
A full householder application means the council assesses the design rather than checking dimensions against a table. That’s slower, with an eight week target determination period running from validation rather than submission. It also means you can design what the house needs rather than what fits the caps.
Plenty of schemes outside permitted development get approved on their merits. The route being longer doesn’t make the extension less achievable.
The Sequence Summarised
| Order | Decision | What it locks in |
| 1 | Plot and planning constraints | What’s possible at all |
| 2 | Width or depth | Extension type |
| 3 | Depth from plot conditions | Footprint and garden retained |
| 4 | Daylight strategy | Roof design and glazing |
| 5 | Structural approach | Beam spans and cost |
| 6 | Threshold and floor level | Slab, drainage, damp detail |
| 7 | Internal layout | Room arrangement |
| 8 | Planning route | Timeline and paperwork |
Run them in that order and each decision narrows the next one sensibly. Run them backwards, starting from a layout, and every earlier decision gets forced to accommodate something that was never tested against the plot.