Home · Business · Sales · Prospecting & outreach
As of 13 August 2026, AI can find prospects using Companies House.
This still needs a person who signs their name to it.
Can you do it?
15 minutesto a draft.
30 minutesto something you’d act on.
Cost, all in£0
Skill neededchat-fluent
Who has to check ityou
What the alternative costsApollo.io is a commercial prospect database with AI outreach sequences and enrichment.
If this goes wrong: you spend time contacting unsuitable or outdated prospects and need to remove them from your list.
What to actually do
Hand it to a person
The route this page recommends
A person who owns the outcome does this end to end, worth it when the failure is dear.
Do it yourself
Second choiceA chat interface, chat-fluent skill, and roughly 30 minutes until you can act on the result.
How to actually do it
- Open Companies House and define the search criteria: target industry or SIC code, UK location, active status, company size and exclusions.
- Gather the relevant Companies House search results and record the company name, company number, status, SIC code and available filing details for each candidate.
- Paste those records into the prompt with a concise description of your product, your ideal customer and the evidence that makes a company a possible fit.
- Ask the model to separate confirmed Companies House facts from inferences and to rank the companies using only your stated criteria.
- Open each cited Companies House record and compare the drafted company number, status, SIC code, registered office area and filing details against the live record.
- Remove companies whose records are inactive, outside your criteria, duplicated or supported only by an unsupported inference.
- Add the checked companies to your CRM and research an appropriate business contact through lawful, current business sources before sending any outreach.
Prompt
Help me find prospect companies from Companies House for a UK sales campaign. Target customer: - Industry or SIC codes: [industry or SIC codes] - Geography: [UK region, town or postcode area] - Company size: [employee range or turnover range if available] - Company status: active companies only - Exclusions: [industries, company types or other exclusions] - What I sell: [product or service] - Signs that a company may fit: [specific criteria] Use only information I provide or information you can directly check in current Companies House records. Do not invent company details, employees, contact information, buying intent or business problems. Separate confirmed facts from your inference. Do not treat incorporation, accounts, charges or director information as proof that a company needs my product. Return a table with: 1. Company name 2. Companies House number 3. Registered office area 4. Company status 5. SIC code and plain-English activity 6. The Companies House record or filing checked 7. Why it may fit my criteria 8. What is still unknown 9. A prospect priority of high, medium or low, with a short reason 10. A safe next research step Do not provide personal mobile numbers or private email addresses. If you identify a director or officer, include them only where the information is publicly shown in the Companies House record and explain why that role may be relevant, without claiming they are the buyer. End with a short checklist of every row I must verify in Companies House before adding it to my CRM or contacting the company.
Open it prefilled in ChatGPT or Claude, or copy it into Gemini, which takes no prefill link.
Use a tool built for this
The distant third
What it gets wrong
- AI cannot infer genuine buying intent from incorporation details, SIC codes or routine filings.
- AI cannot guarantee that a Companies House record is current enough for your campaign or that a company still matches your ideal customer.
- AI cannot choose the right decision-maker merely from a director or officer listing.
- AI cannot transfer responsibility for lawful prospecting, contact selection or outreach to your business.
- AI cannot replace a CRM or enrichment workflow when you need repeatable deduplication, updates and audit trails.
Even on a YES, the friction has a name: real time truth, verification cost and context depth.
How we scored this
Five axes, each scored nought to two by hand: ten means AI carries the task cleanly, and the thresholds that turn a total into YES, PARTLY or NO are published in the methodology. Each axis name links to its definition.
| Axis | Score (0–2) |
|---|---|
| Output | 2 |
| Inputs | 2 |
| Verification | 1 |
| Liability | 2 |
| Effort delta | 2 |
| Total | 9 / 10 |
The methodology and its thresholds are published in full.
FAQ
- Can AI search Companies House for prospects?
- Yes. AI can filter Companies House records, extract company details and produce a prioritised prospect list when you give it clear criteria and the relevant records. It cannot prove that a company has buying intent, so check every row against the current Companies House record.
- Can ChatGPT find companies on Companies House?
- It can help analyse Companies House information and structure a prospecting search, provided the information is supplied or the chatbot has access to current web results. Do not rely on it to invent missing records, contact details or evidence that a company needs your product.
- How do I use Companies House to find sales leads?
- Search for active companies using relevant SIC codes, location and any available filing or size criteria, then export or copy the records you want to examine. Use AI to summarise and rank them, and verify the company number, status and activity before adding anyone to your CRM.
- Can AI find directors' contact details from Companies House?
- AI can identify publicly listed officers and help you find official business contact routes, but Companies House is not proof that a director is the right buyer or that a private contact detail is accurate. Use lawful business sources, avoid private contact data and check that your outreach complies with UK marketing and data protection requirements.
Nearby answers
Assessed by gpt-5.6-luna (gpt-5.6-luna) on 2026-08-13, second-checked by an independent model. Wrong somewhere? Email [email protected] and it gets re-checked.
The newsletter
AI news, new answers and product picks, straight to your inbox.