Ground floor use mix

What a ground floor use mix is, and how one actually gets set

The question. Who decides what ends up on the ground floor, and what can a plan really control?

Ground floor use mix is the phrase for what happens at street level: how much of the frontage is food, how much is retail, how much is a bank or a nail bar or a doctor's surgery, and how much is nothing at all. It is the layer of a building that determines whether a street is somewhere to be or somewhere to leave.

It is also the thing most often specified as an aspiration and least often specified as a number. A design and access statement that promises an active, diverse ground floor has committed to nothing measurable, and will be judged in five years against whatever actually turned up.

There are two separate questions underneath. What is a mix, stated precisely enough to be checked later? And who actually controls it, given that a plan, a landlord and a tenant each hold a different part of the answer?

The second question has a different legal answer in every jurisdiction, and a page that gave one answer would be wrong almost everywhere. So the regime is handled generally here, with the method for finding your own, and the jurisdictions we have written up in full are linked at the end.

The frontages, which are the part of a building the street actually meets.

In one sentence

Ground floor use mix: The set of uses occupying the street level frontages of a place, described as shares by category, and the main lever a plan has over how a street feels and trades.

Written for

A mix is a share, not a wish list

Stated usefully, a mix is a set of category shares over a defined frontage: this many metres of frontage or this many units, split by category, with a date. Fifty-one per cent food and drink across sixty-two units on the date it was counted is a mix. A list of five brands somebody would like is a letting strategy, which is a legitimate document and not this one.

Shares are what make the number portable. A count of eleven cafes means nothing until you know whether it sits in twenty units or two hundred, and it cannot be compared with the centre down the road until it is a proportion. Frontage length and unit count give different answers, because one big unit and four small ones are five units and often the same metres, so say which you used.

The date matters as much as the number. Ground floors churn, and a mix without a date cannot be used to show a change, which is usually the whole reason anyone wanted it.

A frontage survey on foot
Walk the street and photograph every ground floor frontage in order.
What it misses. The only method that catches the unit trading with no online presence and the one still listed but boarded. It is also the only one that sees the difference between an active frontage and a shuttered one.
The VOA rating list
The Valuation Office Agency's public search, covering non-domestic property in England and Wales with a description and rateable value per entry.
What it misses. Gives you a unit list and a rough size without walking, which makes it a good skeleton. It describes the property rather than the current trade, and being a tax record it lags: an entry can describe the last occupier for months.

Every jurisdiction sorts uses into buckets, and the buckets are the lever

Whatever it is called, the regime works the same way in outline. Uses are sorted into categories. Moving a unit between two uses in the same category is usually nobody's business but the landlord's. Moving it across a category boundary needs permission, and that permission is where a plan gets to have an opinion. England calls the categories use classes, most of the United States calls them permitted uses in a zoning district, France sorts them into destinations and sous-destinations, Germany into the use types a Baugebiet allows. The vocabulary differs; the mechanism does not.

Two things follow from that shape, and they hold everywhere. First, the wider a category is, the shorter the lever: a category that contains both a shop and a cafe means a plan cannot stop one becoming the other. Second, the uses held outside the categories, needing their own permission every time, are the uses a policy can still speak to directly. That is why so much local policy anywhere is about hot food, alcohol and gambling: those tend to be the ones held out.

So the useful question about your own jurisdiction is not what the classes are called. It is where the boundaries fall, and which uses were deliberately left outside them. Answer those two and you know the size of the lever you are writing policy with.

The instrument itself, not a summary of it
The statutory text: the Use Classes Order on legislation.gov.uk for England, the municipal zoning ordinance and its use table for a US city, the Code de l'urbanisme for France, the Baunutzungsverordnung for Germany.
What it misses. The only version that is current. Summaries and consultancy explainers lag amendments by years and are the main way an out-of-date category list keeps circulating. Statutory text is hard reading and never explains intent, so read it alongside whatever policy document interprets it locally, and confirm you have the amended version.
The planning or zoning application register
Most planning authorities publish a searchable register of applications and decisions: the local planning authority portal in England, the permit and planning commission record for a US municipality, and the equivalent elsewhere.
What it misses. Shows what has actually been applied for and refused nearby, which tests a policy far better than its wording does. Search interfaces vary enormously, records are often incomplete before some cut-off date, and the changes that needed no permission are invisible here, which is exactly where the mix moved.

What the plan can do, and what the lease does instead

What survives on the policy side is real but narrower than it looks, and the instruments rhyme across jurisdictions. A plan can set frontage and active use requirements on new development. It can resist a change into one of the uses held outside the categories. It can attach conditions to a permission that tie a unit to a use or a size. And most systems have some mechanism for withdrawing an otherwise automatic right within a defined area, so that a change needs an application again: an Article 4 direction in England, an overlay district in much of the US. Find out what yours is called, because it is usually the strongest thing available.

The larger lever, for most streets, is not planning at all. A landlord controls the mix through unit sizes, through permitted use clauses in the lease, through rent-free periods offered selectively, and through who gets a lease at all. A scheme that wants a bakery gets one by making a unit a bakery can afford, not by writing bakery into a policy.

For anyone writing a plan, that division is the useful conclusion. Policy sets the floor and blocks the outcomes nobody wants. The mix that actually appears is decided by unit economics, which means the plan's real influence over mix runs through the sizes and configurations it permits.

Describe an existing mix so it can be compared later

Four things make a mix statement durable. A boundary, so the denominator is fixed. A category list, chosen once and written down, because the difference between a cafe and a restaurant will otherwise be decided differently by each person who surveys it. The unit of measure, frontage metres or unit count, stated. And the date.

Categories are where these exercises quietly fail. Public sources categorise businesses inconsistently and often let the business choose its own label, so the same operator appears as a cafe in one place and a restaurant in another. Whatever list you adopt, the rule that matters is that the same rule was applied at both ends of a comparison, which usually means writing the borderline decisions down rather than trusting anyone to remember them.

The adopted local plan and its monitoring report
Council planning policy pages, usually alongside an annual monitoring or authority monitoring report.
What it misses. Often contains a previous frontage or vacancy count for the same street, which is the only honest baseline you have. The categories used are rarely the ones you would choose, and are rarely documented well enough to reproduce.
Google Maps and OpenStreetMap together
Search categories inside the boundary in one, cross-check the list against the other.
What it misses. The fastest way to a first draft list. Categories are self-assigned and inconsistent in the first, and community-dependent in the second, so both need the walk before the numbers go in a document.

The short version

State a mix as category shares over a defined frontage, with the unit of measure and the date.
Every regime sorts uses into categories. The wider a category, the less a plan can say about a change inside it.
The uses held outside the categories, needing permission every time, are where policy still has a direct say.
Unit sizes and lease terms decide most mixes. A plan's strongest influence over mix is the unit configurations it permits.

Where this goes wrong

Writing the mix as a tenant list

A named wish list dates the moment one of the operators changes strategy, and it cannot be checked. Categories and shares survive a brand leaving the market, and can still be measured in five years, which is when somebody will ask whether the plan worked.

Quoting a category list that has been amended

Use categories get redrawn, and the old names stay in circulation in templates, consultancy reports and submissions for years afterwards. England is the current example: A1, A2, A3, A5 and B1 are still quoted, and any argument built on the old boundary between A1 and A3 is aiming at a line that moved in 2020. Before relying on a category, confirm it against the current statutory text.

Carrying a policy approach across a border

The instruments rhyme between jurisdictions but they do not match, and an approach lifted from another country routinely aims at a lever that does not exist in the same shape locally. The failure is quiet, because the policy still reads well. Check which mechanism actually applies before writing the paragraph.

Comparing two counts made with different categories

The most common way a mix statement misleads is that the earlier count split food and drink differently from the later one. The apparent shift is then an artefact of the category list. Write the category rules down with the count, or accept that the number cannot be used to show a change.

By jurisdiction

The rules differ by country. These are the ones written up in full

Working somewhere not listed? The method in this page holds; the instrument names will not. Ask us and we will tell you what we do and do not know about that market.

Now do it

Or have it run for you

The existing mix, counted the same way at both ends of a comparison

Every business inside a boundary you drew is collected with its category, so the mix is a count rather than an impression.
Categories are applied consistently across runs, which is the condition that makes two counts comparable at all.
Shares are reported rather than raw counts, so a street with sixty units can be set against one with twenty.
The same boundary can be run again later, which is what turns a mix statement into evidence of a change.

What is on the ground floor here, in shares?

Area Analysis

The category mix across the area with shares rather than counts, plus the rating distribution and the themes recurring across those operators.

What would this operator add to the mix?

Tenant Factsheet

One operator read against the area it would sit in: what it is, how its customers talk about it elsewhere, and which part of the mix it fills.

Draw your area and see it

Get started free

More insights