The expensive mistakes at the start of a project are not construction mistakes. They are decisions about arrangement, and every one of them costs nothing to fix on a sketch and a great deal to fix once drawn, surveyed, submitted and priced.
No. A planning drawing must be at a stated metric scale with a north point and a scale bar, and a site plan must sit on a licensed Ordnance Survey base. A sketch meets none of those and is returned as invalid. Sketches belong before the drawings, to test whether an idea is worth drawing at all.
| Stage | What it answers | Accepted by a council? |
|---|---|---|
| Concept sketch | Does this arrangement work at all? | No |
| Measured survey | What shape is the building actually? | Not a submission |
| Pre-application drawing | What would an officer say? | Yes, indicative accepted |
| Planning drawings | May I build this? | Yes |
| Building regulations drawings | How is it built? | Yes, to building control |
Validation requirements are set by each local planning authority under the Town and Country Planning (Development Management Procedure) (England) Order 2015; scale, north point and licensed base are standard across London checklists.
A sketch exists to be wrong cheaply. That is its whole function, and it is why it looks unfinished: the moment a drawing looks resolved, people stop arguing with it.
At the start of a project the expensive mistakes are not construction mistakes. They are decisions about arrangement — putting the kitchen where the light is not, running a stair up through the one place the beam has to go, extending three metres when four would have been permitted anyway. None of those cost anything to fix on a sketch and all of them cost a great deal to fix once drawn, surveyed, submitted and priced.
What a sketch is not is a cheaper planning drawing. It cannot be submitted. Councils require a stated metric scale, a north point, a scale bar and, for site plans, a licensed Ordnance Survey base — and our site plan drawings cover that requirement properly.
Sketching pays when the question is about arrangement. It wastes money when the question is about dimension, because a sketch cannot answer a dimensional question and pretending otherwise produces a confident drawing of something that does not fit.
| Your question | Sketch answers it? | What you actually need |
|---|---|---|
| Should the kitchen go front or back? | Yes | A sketch |
| Can we lose the corridor? | Yes | A sketch |
| Will a 2.4 m island fit with 1 m either side? | No | A measured survey |
| Is there head height for a loft stair? | No | A measured survey and a section |
| Is this permitted development? | No | Drawings and a Lawful Development Certificate |
| What will the council say? | Partly | A sketch plus pre-application advice |
The middle two rows are where projects go wrong. Existing buildings are almost never the shape people believe. A Victorian terrace that reads as rectangular is routinely out of square by 100 mm or more across its width, and a layout sketched against the remembered dimension fails against the real one.
Pre-application advice is the one formal setting where a sketch carries weight. Most London boroughs accept indicative drawings for a pre-app, so a sketch and a written description can buy an officer's written view before you have committed to a full set.
Whether that is worth the fee depends on the borough and on how unusual your proposal is. There is a figure worth knowing here: across London, 7.4% of applications in the twelve quarters to March 2026 were withdrawn before any decision was made. A withdrawal is not a refusal; it is an applicant discovering mid-process that the scheme was not going to work, after paying the £548 householder fee. Borough-level variation on that number is wide, and in the boroughs where it is highest a pre-app is usually the cheaper route.
Source: MHCLG District Planning Application Statistics (PS1 and PS2), decisions made in the twelve quarters 2023 Q2 to 2026 Q1, retrieved 1 September 2026. Open Government Licence v3.0.
The sketch settles the arrangement. Everything after it makes the arrangement real, in a fixed order, because each stage depends on the one before.
| Stage | What it produces | Why it comes here |
|---|---|---|
| 1. Concept sketch | An agreed arrangement | Cheapest place to change your mind |
| 2. Measured survey | The building as it actually is, with levels | Everything downstream is drawn against it |
| 3. Planning drawings | Plans, elevations, sections, site plans | Establishes what you are permitted to build |
| 4. Building regulations drawings | Details, annotated sections, specifications | Drawn against the permission actually granted |
| 5. Structural design | Beam sizes and calculations | Confirms what the details were drawn assuming |
Skipping step two is the single most common sequencing error, and it is expensive because it invalidates steps three, four and five at once. Our measured building survey is included in every package for that reason, and the planning drawings are never begun before it.
It cannot be built from, priced reliably, or submitted anywhere that matters. A sketch carries proportion and intent; it does not carry dimensions anyone should cut timber to. That is not a shortcoming to be fixed by drawing it more neatly — it is the point. A sketch is cheap precisely because it leaves out everything that has to be verified, and the moment you add those things you are paying for a different drawing.
Three things go wrong when a sketch is asked to do more than it can. A builder prices from it and the price is fiction, because nothing on the sheet fixes a wall position. A client signs it off believing the window is where it appears, and discovers at the survey that a soil stack occupies that corner. And a sketch gets sent to a council, which is not a submission, produces no decision, and buys nothing but a polite reply.
Because it is drawn on assumptions about a building nobody has measured. Estate agent floorplans are indicative and frequently wrong by 200mm or more per room, which is enough to move a staircase from workable to not. Older London stock is worse: walls are rarely parallel, floors step between the back addition and the main house, and a Victorian rear return is almost never the width the deeds imply.
So the honest sequence is survey first, sketch second, when the stakes are real. If you are deciding whether to buy the house at all, sketch on what you have and accept the error bar. If you are deciding where the kitchen goes, measure first — a sketch drawn on a measured building survey costs the same to produce and is worth acting on. The difference is entirely in what sits underneath it.
There is one exception worth naming. If the question is purely about massing — will a two-storey addition here overshadow the neighbour, does a mansard read badly from the street — a sketch on an approximate survey answers it perfectly well, because the error bar is smaller than the thing being judged. It is interior planning, where 200mm decides whether a door swings, that punishes a guess.
Answer four questions and a competent designer has enough. What is the problem — not the solution you have in mind, but the thing that is wrong now: no room to eat, dark at the back, nowhere to put a fourth bed. What must not change, which is usually a budget ceiling, a tree, a party wall, or a room someone refuses to lose. Who lives here in five years, since a fourth bedroom and a bigger kitchen are different buildings. And what have you already been told, by a neighbour, a builder or a planning officer, whether or not it turned out to be true.
What is not useful in a brief is a floorplan copied from a house you liked online. It arrives as a fixed answer to a question your site has not been asked, and most of the ten minutes then goes on unpicking it. Bring the photograph of the room you liked instead; that carries the intent without smuggling in a plan that does not fit.
A quick, deliberately rough drawing used to test whether an idea works before anyone pays for measured drawings. It is not a planning drawing and it is not to scale in any way a council would accept. Its job is to answer 'is this worth pursuing' for a fraction of the cost of finding out later.
No. Planning drawings must be to a stated metric scale, on a licensed base where a site plan is involved, with a north point and a scale bar. A sketch fails all of those, and a submission built on one is returned as invalid before an officer reads it.
It depends what you are unsure about. If the question is 'will this layout work', a sketch answers it cheaply. If the question is 'will it fit', only a measured survey answers it, because the existing building is almost never the shape people think it is.
No. Concept work is part of the design stage inside every package from £840. We would rather spend an hour establishing that an idea does not work than draw it properly and have you find out on site.
A sketch tests whether an arrangement works. A visualisation shows what it will look like once it does. They belong at opposite ends of the process, and paying for a photorealistic image of a layout nobody has stress-tested is the most common way to waste money early.
Yes, and it is one of the few places a sketch has real standing. Most London councils accept indicative drawings for pre-application advice, so a sketch plus a written description can buy you an officer's view before you commit to a full set.
Site plans are included in every drawing package we produce, on a licensed Ordnance Survey base with the ownership edging done properly.
Get a quote
Before you commit, it is worth understanding what it costs and seeing why homeowners choose us. You can search the site to find the right fit, or open a client account to track your project. When you are ready, get a quote today — most homeowners have a fee in minutes — or contact us about your project and a chartered technologist will be in touch.
Each area page carries the local planning authority, conservation-area position, postcode districts and typical housing stock for that address.
