Free tool · no sign-up · no email
How crowded, and how healthy, is a UK registered office address? Graded A–E from Companies House data, and we grade our own address on the same page.
Companies House data · checked 25 August 2026
The tool
Type any UK postcode or address, or run one of the worked examples. Six postcodes are indexed so far and every figure in them is a real count taken from the Companies House public register on 25 August 2026 – nothing on this page is illustrative. The league table further down lists them all, and every one is clickable.
Check any UK company address free. No sign-up, no email.
Counts come from the Companies House register. Our index is rebuilt monthly.
Design decisions
01
Not a letter in a circle. One square per company, coloured by status. At a healthy address you see a handful of squares; at a formation address, a mat running off the edge. No adjective is used – it is literally the data, and it is the most screenshot-able thing on the page.
02
No sign-up and no email. A gate kills sharing, kills inbound links and kills the press value. We ask for an email only where something genuinely additional is on offer: the written report, monitoring, or a transfer quote.
03
“25,736 companies are registered here” is reportable. “Your bank will reject you” is not. Every outcome is described as what providers commonly do, never as what will happen, and no provider is ever named as making a decision.
04
A healthy address is told it is healthy and offered nothing to buy. A tool that only ever finds a problem is a sales form wearing a diagnostic’s clothes, and that is exactly what a journalist would say about it.
05
Both London examples score full marks on liquidation and administration while still grading E overall, and the Leeds example loses marks on dissolution rate and still grades A. A model that always agrees with the sales pitch is not a model.
06
Filing health needs the Companies House bulk file, which our index does not yet hold. So that factor reads “not yet indexed”, its weight is spread across the others, and the grade is marked provisional. An honest gap beats an invented number.
The league table
Every figure below is a count from the free public Companies House search, taken on 25 August 2026. We name postcodes, never companies. What is registered where is a matter of public record; what you make of it is your business.
| Postcode | What is registered there | On the register today |
|---|---|---|
| WC2H 9JQ | Covent Garden, London | 87,314 |
| N1 7GU | Hoxton, London | 36,384 |
| EC1V 2NX | Old Street area, London | 25,736 |
| EC2A 4NA | Shoreditch, London | 9,990 ever registered |
| E17 3NU | 195-197 Wood Street, Walthamstow | 3,798 |
| RM6 6AX | 321-323 High Road, Chadwell Heath | 2,872 |
| LS11 5AD | Great Wilson Street, Leeds – a genuine head office, for scale | 48 |
At WC2H 9JQ, 104,705 companies have been dissolved. That is more than are currently on the register at the same address.
Postcode is the unit because Companies House stores addresses as free text and the postcode is the only boundary that compares reliably. Some of these are a single building – every company at WC2H 9JQ sits in one building – and some take in neighbours, which is why EC1V 2NX and EC2A 4NA are described by area rather than by building. Counts run against the free public advanced company search, so anyone can repeat them. Click a postcode to run it through the checker above. Contains public sector information licensed under the Open Government Licence v3.0.
Methodology
A tool that grades other people’s addresses has to be willing to show its working. Here is every factor, every weight and every band, so you can check our arithmetic against the register yourself.
| Factor | Weight | What it measures | Bands |
|---|---|---|---|
| Companies registered here now | 25% | How many share the address today. A handful is a real office; thousands is a formation address, and that is the difference a bank reacts to. | 1–49 = 100 · 50–199 = 80 · 200–999 = 55 · 1,000–4,999 = 25 · 5,000+ = 0 |
| Dissolution rate | 20% | Dissolved as a share of every company ever registered at the address. Companies come and go everywhere; when most of them have gone, that is a pattern. | up to 10% = 100 · 10–20% = 75 · 20–30% = 50 · 30–45% = 25 · over 45% = 0 |
| Liquidation and administration | 15% | Live companies at the address currently in liquidation or administration, as a share of the live count. | up to 0.5% = 100 · 0.5–1% = 75 · 1–2% = 50 · 2–4% = 25 · over 4% = 0 |
| Filing health | 15% | The share of companies there with accounts or a confirmation statement overdue. | Not yet indexed – weight redistributed, grade marked provisional |
| Formation velocity | 15% | New incorporations in the last 90 days measured against the address itself. A burst at a quiet address is a formation run. | up to 1% = 100 · 1–3% = 75 · 3–6% = 50 · 6–12% = 25 · over 12% = 0 |
| Company age profile | 10% | The share of live companies incorporated in the last three years. A very young profile means churn. | up to 20% = 100 · 20–35% = 75 · 35–50% = 50 · 50–70% = 25 · over 70% = 0 |
| Score | Grade | Reading |
|---|---|---|
| 90–100 | A | Low density, healthy register |
| 75–89 | B | Normal for a shared business address |
| 60–74 | C | Busy – expect additional checks from some providers |
| 40–59 | D | High density with elevated dissolution or filing failure |
| 0–39 | E | Mass-formation profile |
Below thirty companies ever registered, a percentage is arithmetic rather than evidence – one closure out of four is 25%, and it means nothing. At those volumes we score on density alone, show every raw count, and say so on the result. Take an address where thirteen of the twenty-one companies ever registered have closed: at a formation address that ratio would be a serious flag, and at a small office with a long history it means almost nothing at all.
A well-run business centre with 400 healthy companies is a safer address than a letterbox with 400 of which half have been struck off. That is why density carries a quarter of the grade and not all of it – and why the London example on this page scores full marks on one factor even while grading E. A model that always agrees with the sales pitch is not a model.
Companies House stores addresses as free text, so the only boundary that is reliably comparable is the postcode. A postcode usually covers one building – every company counted at WC2H 9JQ sits in a single building – but it can cover several; EC1V 2NX takes in more than one. So we report the postcode and describe the area, and we never present a postcode count as the count for one company’s office. Building-level attribution needs the full Companies House bulk file and the address normalisation described below; until that is running, postcode is the honest unit.
The full model runs against the Companies House monthly bulk file. That pipeline is still being built, so today the tool holds a small number of addresses that we have counted by hand against the public register, and every one of them shows the date it was checked. Ask it about an address we have not reached yet and it will say so, point you at the official Companies House count for that address, and offer to run the whole check by hand and email it to you free. What it will never do is invent a grade to fill the gap.
Companies House addresses are free text, so “The Exchange” and “The Exchange” are the same building written two ways. We strip unit, floor and PO box prefixes, standardise street types and roll the rest up. It is good, not perfect, so results offer two scopes: this exact address, and this postcode. The postcode is always reliable; the address is more precise.
We will not tell you an address is blacklisted, flagged by HMRC or fraudulent, because we cannot know any of those things and neither can anybody else selling you an address. We will not put words in a bank’s mouth. We will not suggest anything about the intentions of an address operator or of the companies registered there. We give you the count, the date and the source, and you draw your own conclusion.
Data source: Companies House, via the free public advanced company search that anyone can run. Figures on this page checked 25 August 2026. Contains public sector information licensed under the Open Government Licence v3.0.
Cookies. This site sets only the essential cookies it needs to work. With your permission we also use analytics cookies to understand how the site is used – nothing else, and never for advertising. Cookie policy