Why a site visit cannot be skipped
Datasets give you flood zones and listing entries. Levels, boundary positions, neighbouring window heights and visibility splays come from standing on the plot with a tape measure.
Aerial imagery will not tell you the level difference across a plot, where a neighbour's bathroom window sits, or whether the gatepost is two metres from where the title plan says the boundary runs. Those three facts have sunk more applications than any policy argument. A site visit planning report exists because a category of evidence only exists on the ground, and no dataset, satellite photograph or AI-drafted document can substitute for someone standing on the plot with a tape measure and a camera.
Self-builders feel this most acutely, because the cost of a measured survey looks avoidable right up to the point a council returns the application or a neighbour produces a photograph you have never seen.
Facts that only exist on site
Three sorts of information behave differently.
Information held in datasets can be looked up: flood zones, listed building entries, conservation area boundaries, tree preservation orders, ward and parish boundaries, soil classification, geology. Automatic constraint checking handles this reliably and there is no reason to walk a site to find out whether it is in Flood Zone 2.
Information held in documents can be requested: planning history, approved drawings for neighbouring schemes, title plans, deeds of easement, service records from utility providers.
Then there is information that exists only as a physical condition of the land on the day you look. Levels, actual boundary positions, the state of a wall, which windows face you, where the water sits after rain, how fast cars come round the bend. This is the category that cannot be inferred, and it is the category that decides applications.
The honest version of this distinction is the subject of what AI drafts get right and wrong in planning reports, and the dividing line is almost exactly the one above.
Levels, boundaries and encroachment
Levels first, because they propagate into everything. A 600 millimetre fall across a plot changes the ridge height of a proposal relative to its neighbours, changes whether an extension needs steps or a ramp, changes the drainage design and changes the daylight result for the house next door.
A topographic survey tied to a datum costs a fraction of what a redesign costs. Where a full survey is not proportionate, take spot levels at the corners of the plot and at the threshold of every adjoining building, and record the datum you used.
Boundaries are the second recurring problem. And they are worse, because nobody expects them. The Land Registry title plan shows general boundaries, not precise lines, and the physical boundary on the ground is frequently somewhere else. A fence erected 300 millimetres inside a boundary 40 years ago, a wall rebuilt on a different line, a strip of land maintained by a neighbour without title: each of these turns a straightforward application into a dispute. Walk the perimeter, photograph every boundary feature against a scale, and note where the physical line and the title plan disagree.
Encroachment runs both ways. Scaffolding, eaves, gutters and foundations crossing a boundary need the neighbour's agreement regardless of planning permission, and an ownership certificate signed on a false assumption about who owns the land is a defect in the application itself.
Neighbour windows and outlook
Daylight, overlooking and overbearing effect are all assessed from positions you can only establish by looking.
Record, for every neighbouring facade that faces the site: the position of each window, its head and cill height above external ground level, which room it serves if that can be established without intrusion, and whether it is the only window serving that room. A single aspect room behind one window is treated very differently from a dual aspect room.
Also record what is not a window. Obscure glazed bathroom windows, rooflights, secondary windows and non-habitable rooms change the assessment, and councils have refused schemes on the basis of a window that turned out to serve a landing.
Outlook works the other way. Stand where the proposed windows will be and look. A first floor rear window that stares into a bedroom eight metres away will attract an objection that no amount of policy drafting answers, and the fix, an obscure glazed panel or a repositioned opening, costs nothing at design stage.
Trees, drains and services
Trees on and near a site are assessed under the British Standard for trees in relation to design, demolition and construction, which requires stem diameter measured at 1.5 metres, crown spread in four directions, height and an assessment of condition. None of that comes from aerial photography, and the root protection area calculated from stem diameter is what determines whether a foundation position works.
Trees on neighbouring land matter as much as trees on the site, because their root protection areas cross the boundary and their canopies create the shading.
Drains and services need a different kind of looking. Lift the inspection chamber covers. Note the cover levels, invert levels, the direction of flow and whether the run is combined or separate. A shared drain crossing the proposed footprint needs either a diversion agreed with the sewerage undertaker or a build-over agreement, and discovering it during excavation is expensive.
Overhead cables, street lighting columns, telecoms cabinets, gas governors and electricity substations all constrain the design and all are visible on a walk round.
Access geometry and visibility
Visibility splays are the single most common reason a highway authority objects, and they depend on measured facts.
What to record at the access:
- The width of the existing access and the kerb radii.
- Available visibility in both directions, measured from a set-back point, with the obstruction that limits it identified: a hedge, a parked car, a wall, a bend.
- The actual speed of traffic, because splay requirements are set from recorded speeds rather than from the posted limit, and on many rural lanes the recorded speed is well below the limit.
- Footway width and continuity, and whether there is a dropped kerb or a crossing nearby.
- Where refuse vehicles currently turn and whether they can reach the proposed bin store.
Photographs taken from driver eye height, with a scale in frame, are worth more than a statement. On a constrained access, a swept path drawing done from measured dimensions will either demonstrate the access works or tell you early that it does not.
Condition of existing fabric
For a conversion or an extension to an older building, the fabric is evidence.
Look for movement cracking and its pattern, damp staining at low level, the condition of the roof structure where it can be seen, whether the existing foundations are likely to be shallow Victorian footings, and what the building is actually built of behind the render. A Structural Survey commissioned before the application rather than after permission tells you whether the design is buildable, and on a barn or a chapel conversion the council will want to know the building can be converted without substantial rebuilding.
Materials matter for the design case. Take close photographs of brick bonds, mortar joints, window details, eaves and rainwater goods. A design statement that specifies matching materials based on a photograph of the actual wall reads differently from one that specifies red brick.
Recording a visit so it is useful
A visit nobody can reconstruct is half a visit. The discipline that makes it count:
- Date and time, weather, and who attended. Time of day matters for shadow and traffic observations.
- Photographs numbered and keyed to positions marked on a site plan, with the direction of view shown. Unkeyed photographs are the most common failing.
- A sketch with dimensions taken on site, not redrawn later from memory.
- A written list of assumptions you still hold and what would confirm each one, which becomes the list of items flagged as unconfirmed in the draft documents.
- A note of anything a neighbour said, with the date, since it often predicts the objection.
PlanScope will extract coordinates from a site photograph where the image carries them and check the site against flood, heritage and designation data automatically, which handles the dataset half of the job. The drafts insert flags where site-specific facts are missing rather than inventing them, and those flags are the list of things the visit has to answer. Reports should be reviewed before submission and PlanScope does not guarantee planning approval.
Where the measurements need a professional, use one. The reviewers in the professional directory can take a draft and the site record and provide the sign-off, and for specialist technical reports qualified professional input may be required. A visit of two hours with a tape, a camera and a drain key removes most of the reasons applications come back, which is a shorter list than why planning applications are returned as invalid suggests but overlaps it heavily.
This is general guidance, not legal advice; check the position with your local planning authority or a qualified professional before relying on it.
Frequently asked questions
What can be checked from datasets without visiting a site?
Flood zones, listed building entries, conservation area and Green Belt boundaries, tree preservation orders, soil classification, geology and designation layers are all held in published datasets and can be checked reliably from a postcode. Planning history, approved neighbouring drawings and title plans are documents that can be requested rather than observed.
Why do levels matter so much on a planning application?
A fall across a plot changes ridge height relative to neighbours, whether steps or a ramp are needed, the drainage design and the daylight result next door. A topographic survey tied to a datum costs a fraction of a redesign. Where a full survey is disproportionate, take spot levels at plot corners and adjoining thresholds.
Is the Land Registry title plan enough to establish a boundary?
No. Title plans show general boundaries rather than precise lines, and the physical boundary on the ground is often somewhere else after decades of fences being rebuilt. Walk the perimeter, photograph every boundary feature against a scale, and record where the physical line and the title plan disagree before signing an ownership certificate.
What should be recorded about neighbouring windows?
The position of each window on every facade facing the site, head and cill heights above external ground level, which room it serves where that can be established without intrusion, and whether it is the only window to that room. Note obscure glazed, secondary and non-habitable windows too, because they change the assessment.
How should a site visit be recorded?
Date, time, weather and attendees, photographs numbered and keyed to marked positions on a site plan with the direction of view shown, a dimensioned sketch made on site rather than redrawn later, and a written list of remaining assumptions. Note anything a neighbour said, since it often predicts the objection.
Sources and further reading
This article is general guidance about the planning system in England and, where stated, the rest of the UK. It is not legal or professional advice. Check your local planning authority’s validation requirements and take professional advice for your specific site.