What should company monitoring actually do?
Company monitoring compares a company’s available records over time so your team can identify changes that deserve review. The useful output is a specific task: confirm a new director, investigate a changed shareholder or review a newly filed set of accounts. A stream of alerts without ownership creates another inbox.
Start with the business decision. Procurement may need to know whether the contracting entity has changed. A company verification team may need to reopen an ownership review. A data team may need to propose an update to a registered address. The same event can produce different work for each team.
1. Establish the company and its baseline
Use country and registration number, alongside the legal name, to resolve each company before monitoring it. Preserve the internal supplier, account or counterparty ID so a change can be routed back to the right business record. Treat a similar name as a candidate match until the identifiers are checked.
The first successful retrieval establishes a baseline. It does not demonstrate that the company has just changed. Store the fields retrieved, unavailable datasets, source references and retrieval time. A later comparison must use compatible records for the same entity.
2. Decide which changes should open a review
| Recorded change | Question for the reviewer | Suggested owner |
|---|---|---|
| Registered name or status | Is our legal-entity record still correct? | Company verification team |
| Registered address | Does the registry address need updating? | Data steward |
| Director appointment or resignation | Is the change relevant to the relationship? | Account or supplier owner |
| Shareholder or group relationship | Does our existing ownership assessment need revisiting? | Ownership reviewer |
| New financial filing | Are the periods and reported figures comparable? | Financial analyst |
Document thresholds and exclusions before launch. You might route every recorded status change for review, but group routine address-format corrections into a maintenance queue. Choose policies that your team can explain and maintain.
3. Separate when it happened from when you saw it
Keep effective, filing and detection dates distinct. A change can be filed after it takes effect and detected after the filing becomes available. “Checked today” describes a retrieval; it does not prove that every value is current today.
Set the checking cadence around the decision and source availability. Some portfolios justify frequent checks, while others need periodic review. A schedule is useful only if failed checks are visible and someone owns recovery. An unavailable response must never be treated as evidence that nothing changed.
4. Build the evidence packet and handoff
- Identify the affected record. Include the company, internal reference and changed field.
- Show both values. Keep the prior record and the new record accessible.
- Explain the finding. Summarise the observed difference and attach supporting source references.
- State the open questions. Missing ownership layers or source dates should remain visible.
- Assign the next step. Give the reviewer a task, due date and clear decision options.
- Record the outcome. Preserve the reviewer’s rationale and any approved update.
5. Measure a representative pilot
Include companies from your main countries, several record types and known exceptions. A pilot containing only easily matched domestic companies will hide the work required for your wider portfolio.
| Measure | How to calculate it |
|---|---|
| Match rate | Confirmed company matches ÷ companies submitted |
| Usable field coverage | Matched companies with the required field ÷ matched companies |
| Review usefulness | Reviewed findings that required action ÷ findings reviewed |
| Preparation time | Time spent collecting and comparing evidence for each review |
| Unresolved checks | Failed or incomplete checks still awaiting follow-up |
Choose acceptance criteria before you see the results. A useful pilot demonstrates that reviewers can understand, investigate and close the work, rather than simply producing a large alert count.
How this maps to CompanyDelta
CompanyDelta brings company lists, saved records and investigation reviews into one workspace. Configure the company-data connection, confirm matches and establish successful baselines before relying on comparisons. Connected AI can prepare findings from captured evidence for human review.
Recurring checks need a configured schedule. External delivery also needs an active connection and appropriate access. The website examples illustrate the workflow; they do not demonstrate an active monitoring service for your companies.
Use the monitoring pilot template to document scope and acceptance criteria.
Questions about company monitoring
Is ongoing monitoring the same as ongoing due diligence?
Monitoring identifies changes in the information being checked. Ongoing due diligence is a broader review process that decides what those changes mean and what further work is required. A company-record alert is an input to that process.
What happens when a check fails?
Keep the last successful record, record the failed attempt and route unresolved failures for follow-up. Do not replace a known value with an empty field or label the company unchanged because retrieval failed.
Can every company field be monitored?
No. Availability, source updates and historical records differ. Confirm the required fields against a representative company list before choosing the monitoring scope.