Why I built CompanyIQ

I built CompanyIQ to solve a problem that I have seen and experienced across my career. Companies invest time and resources to assess and qualify companies they deal with, using Companies House, third-party websites, company websites, Google searches, and whatever else they can find. The problem is that too often the results vary by person, even if the steps are the same.

Across the industries I have worked in, there is no consistent or concise process that anyone can agree on. I have worked with sales, audit, and finance teams, and none of them used the same checks, and all of them disagreed about which types of companies are good to work with or which should be avoided. It often led to financial risk or future bad debt, and a lot of the signals or risks could have been identified if they had a single process that worked and provided an objective standpoint, rather than a disjointed opinion.

What 15 years in data work taught me

I have spent most of my career in data, business intelligence and leadership roles. I have worked in retail, telecoms, and other sectors in between. All the companies I worked at made decisions about other organisations, and in every one of those contexts, the same problem showed up in different guises.

The sales team had a qualification framework on paper, but it was applied differently depending on who was running the process. The accounts department had a standard checklist, but each analyst weighted the items according to their own judgment. I have managed dedicated data enrichment teams with a strictly detailed outline that standardised a process. The result was inconsistent at best. The process itself was robust, but there was still too much interpretation and opinion because the source material is varied, not easy to understand, and ultimately, we all interpret things differently.

This is not a criticism of any of those teams or processes. It is an observation about what happens when interpretive work is designed to scale and be consistent. After watching it happen often enough, I started to think about what it might take to do this consistently, and apply it the same way, objectively, every time.

The data has always been there

Companies House is one of the most comprehensive public company registers in the world. The strict rules ensure that every limited company:

• Files annual accounts.
• Records director appointments and resignations.
• Dates every filing event.
• Publishes every charge against a company's assets.

The data is available, but that is not the issue. The issue is how to access and read it. Even Companies House takes multiple tabs, or constant switching between them, to access it all. Most existing tools in this space show you what is filed. They present the data in more or less structured ways, but rarely identify what the data means or why it is important. Businesses either accept inconsistent qualification work and audit findings, or pay for traditional credit reports that do not help them understand the data.

The opportunity I saw was a tool that applies the same interpretive logic to every company and presents actual analysis. CompanyIQ is not aiming to replace trained financial analysts, but scaling a single analyst or team's output is difficult to achieve, and impossible to achieve with a consistent output. CompanyIQ provides an analyst's level of insight for companies that need scale and speed, or do not have a process at all.

What CompanyIQ does

You enter a UK company name or number. CompanyIQ retrieves the data from Companies House, reads the filed financial documents using AI, and returns a scored intelligence report in two to four minutes.

The report covers five dimensions:

• Financial Health measures revenue trajectory, profitability, liquidity, and debt burden, drawn directly from the filed accounts.
• Filing Compliance counts late and overdue filings against Companies House deadlines.
• Director Quality looks at average tenure, recent turnover, and any links to dissolved or insolvent companies.
• Market Position considers company maturity, employee growth, and sector context.
• Risk Assessment surfaces specific concerns supported by evidence from the documents, with severity rated.

These five dimensions combine into a single CIQ Score between zero and one hundred. The methodology is published, and every finding cites the section of the accounts or filing event that supports it. I want the platform to be transparent, not secretive. The data is all available, and there are no subjective data points in the process.

In practice this means a sales team qualifying prospects gets the same answer regardless of who runs the analysis. A commercial landlord vetting a prospective tenant does not need to bring in a financial analyst. A credit team can stop debating whether the analysis is right and start debating what to do about it.

The same data, the same method, the same answer, every time.

What has been hard

I have built CompanyIQ around a full-time job, alongside a young family, with my wife running her own business in parallel. Ten to fifteen hours a week, sustained over many months. A few things have been harder than I expected.

Reading filed accounts consistently was harder than I thought. UK accounts come in wildly different formats. Some are clean, modern PDFs. Some are scanned documents that need OCR before they are machine-readable. Some are short receipts-and-payments accounts for micro-entities. Some are full SORP-compliant accounts running to thirty pages. Getting reliable financial extraction across all of that took real engineering work.

Building infrastructure that handles concurrent users without breaking was harder than I thought. The pipeline is not a "send to AI, get a response back" operation. It is a multi-stage asynchronous system with retries, stuck-job recovery, credit refunds on failure, and observability across every step. Most of what I have learned about production engineering in the last six months has come from running this thing at small scale and watching what breaks.

Resisting the urge to keep tinkering has been the hardest discipline of all. The work I find satisfying as a builder, code, architecture, infrastructure, is not the work that determines whether CompanyIQ succeeds. The work that determines that is the work I find harder: posting, reaching out, having conversations, writing this blog post. I knew this intellectually before I started, but only now am I learning to confront that head on.

What I want from CompanyIQ now

I am not raising money. I am not trying to disrupt anything. I am looking for the right people to try CompanyIQ and tell me whether it actually solves the problem I have described.

If you make repeated decisions about UK companies, in any of the following contexts, I would like to know what you think:

• A sales team qualifying prospects before investing in a sales cycle.
• A commercial landlord assessing prospective tenants.
• A credit team deciding whether to extend terms, and on what conditions.
• An accountant or advisor onboarding new clients.
• A procurement team vetting suppliers or contractors.
• A compliance or risk team running background checks on counterparties.

All first analyses are free, no card required. Try CompanyIQ on a company you already know, your employer, a customer, a supplier, and see whether the analysis matches your instinct or process. If it does, that tells you something. If it does not, that tells you something more interesting.

My email is simon@company-iq.co.uk. I read everything that comes in. I am building this carefully, and all feedback will shape what comes next.

If you have felt the problem I am describing, I would love to know whether CompanyIQ feels like the answer to you.