Nominations now open — Submit before 15 December 2026 · Free to enterSubmit Now

Records Retention Becomes a Corporate Services Evidence Problem

★
Editorial Team
DATENov 22, 2026
Records Retention Becomes a Corporate Services Evidence Problem

Records retention has traditionally been managed as housekeeping. A schedule is drafted, a policy is signed off, an archive is maintained and a destruction date is set. The question the firm asked itself was narrow: have we kept this long enough? That question no longer describes the exposure. For corporate services providers holding entity, payroll, accounting and compliance records, retention has become an evidence problem. The harder questions are why a given record is held at all, what would suspend its deletion, and how the firm can demonstrate that a deletion actually happened across every system that held a copy. The 2026 Global Corporate Services Awards category is a useful reference point for that shift, because clients, banks and auditors are increasingly asking providers to describe retention as an operating control rather than a filing convention.

Retention Is Not One Schedule but Several Overlapping Ones

Corporate services firms sit at the intersection of obligations that were written by different regulators for different purposes. Company law requires statutory registers, minutes and resolutions to be maintained and available. Tax authorities set their own periods for accounting records and supporting documentation. Anti-money-laundering regimes require client due diligence material to be retained for a period after the relationship ends, so that a supervisor can reconstruct what the firm knew and when. Employment and payroll rules add local requirements tied to the worker rather than the client. Privacy law adds the opposite pressure: data should not be kept longer than necessary, and individuals may ask for erasure.

These obligations do not reconcile neatly. A privacy-driven deletion request can collide with a tax retention requirement and an AML record-keeping duty. A client instruction to destroy files at exit may be overridden by a statutory obligation to keep a copy. Litigation or a regulatory enquiry can freeze everything.

The practical consequence is that a single group-wide retention schedule is rarely accurate. A provider needs to be able to state, for each record class and each jurisdiction, which obligation drives the retention period and which one limits it. Where those answers conflict, the conflict should be documented and resolved deliberately, not left to whichever system happens to delete first.

Statutory Records and Operational Data Need Different Rules

The most useful distinction is between records the firm is obliged to hold and data it has accumulated because holding it was easier than deciding what to do with it. Statutory and entity records — registers, constitutional documents, filings, resolutions, ownership evidence — usually have a clear legal basis and a long or indefinite life. Operational personal data is different: onboarding working files, email threads, draft payroll schedules, helpdesk tickets, meeting notes and duplicated client documents held in collaboration tools.

That second category tends to grow without a decision. It is often the larger privacy exposure and the harder one to search when a client asks what the firm holds. Assigning ownership by record class, limiting access to those who need it, and applying separate retention clocks to statutory and operational material makes both obligations easier to evidence.

Keeping everything indefinitely is not a neutral position. It is a decision not to decide, and it enlarges the set of records that must be searched, disclosed, migrated or explained after a breach. A defensible programme can say what it does not hold, and why.

Legal Holds Must Suspend Deletion Reliably

A legal hold that exists in a policy but not in the systems is not a hold. Retention rules are increasingly automated, and automated deletion runs on schedule whether or not someone has remembered to flag a matter. The control only works if the suspension reaches the deletion jobs themselves, including those inside software the firm does not operate.

A workable hold process has several components: identification of the custodians, systems and record classes affected; a suspension instruction that propagates to retention rules, archive jobs and third-party platforms; documented approval by someone with authority to impose it; periodic review while the hold remains in force; and a documented release that is as controlled as the imposition.

Testing matters more than documentation. A firm can select a held record set, allow a scheduled deletion cycle to run, and confirm that nothing was destroyed. It can also confirm that release restores the original retention clock rather than leaving records in indefinite suspension. Holds that are never tested tend to fail at the point they are most needed.

Deletion Evidence Has to Reach the Copies

Deleting a record from the primary system is the straightforward part. The difficulty lies in the places copies accumulate: backup generations held on a rolling cycle, email and collaboration archives, version histories, recycle bins and soft-delete areas in software-as-a-service platforms, and the systems of vendors and sub-processors who process client data.

Evidence should show what was deleted, from which system, on what basis, by which process or person, and when. Where a vendor holds a copy, the firm needs a confirmation it can rely on rather than an assumption. Backup handling deserves specific attention, because deletion from production does not remove backup copies, and a restore can reintroduce records that were properly destroyed. Firms that have tested a restore and then re-applied deletion to the restored data know something about their environment that a policy document cannot tell them.

This is not an argument for deleting more. Records that must be kept should be kept, and the evidence should distinguish clearly between records retained under an obligation and records deleted because no basis to hold them remained.

Client Exit and Due Diligence Ask for Statements That Match Behaviour

Retention now appears in client questionnaires, audit requests and due diligence packs, usually alongside questions about sub-processors, data location and deletion capability. The risk is an answer that describes the policy rather than the system. A stated retention period that the platform does not enforce, or a deletion commitment that stops at the primary application, is a discrepancy a client's auditor may find before the provider does.

Exit is where the gap becomes visible. A client leaving expects its records returned, copies retained only where an obligation requires it, and the remainder deleted within a defined period with confirmation. That outcome depends on the firm knowing which records fall into which category before the exit begins.

The test is consistency. The policy, the system configuration and the answer given to the client should tell the same story, and the firm should be able to demonstrate the third using evidence from the first two.

Conclusion

Records retention is moving from an administrative function to a client-facing evidence and liability issue. The firms that manage it well will be able to answer three questions for every record class they hold: why it is kept, what would suspend its deletion, and how a deletion was evidenced across systems, backups and vendor copies. Those answers are becoming part of how corporate services providers are assessed, and the 2026 Global Corporate Services Awards category reflects a market in which retention statements are read as commitments rather than formalities.