Corporate communications service standards: six measures defined by event, not aspiration

Corporate communications service standards are useful only when a buyer can tell whether a defined event passed. This article supplies a buyer-owned standard template to make that test explicit. An England comms team would populate it with its own thresholds after checking risk, affected people, channels and contractual evidence.

What to take away

Specify the measure, not the aspiration

For each standard, write:

One worked row shows the level of detail required:

A completion rate, for example, is meaningless without an eligible population and a definition of completion. A response duration needs a timestamp source, pause rule and timezone. Do not turn an average from unlike messages into a target.

The NCSC's logging guidance recommends deciding which questions logs must answer and identifying where records arise, how long they are retained and how they are protected. For a corporate publishing standard, those points belong in the Source, query and version field.

Build standards around controllable events

An England comms team could define measures for source-copy completeness, claim-evidence review, accessibility acceptance, approved-version release, correction acknowledgement, controlled-copy replacement and archive closure. Each measure ends at a buyer-controlled state. The team must not claim that an employee understood the proposal, a journalist accepted a view or the public trusted the organisation unless separate research supports that conclusion.

Write the acceptance event clearly before expressing a rate. For source completeness, the eligible population might be the claims admitted to one version, while the numerator is claims with a current evidence ID and authorised owner. For controlled-copy replacement, eligibility might cover destinations the buyer can change. Third-party discussion outside that control belongs in monitoring, not the denominator.

Accuracy needs a challenge path as well as a target. The ICO's accuracy guidance addresses personal information, source status and correction; the ICO marks it as under review after the Data (Use and Access) Act 2025, so check the current version. A personal-data correction and a correction to general corporate copy are different events, even if one incident triggers both. That difference belongs in the Eligible items or people field.

Accessibility should be tested against a named task and format. Government accessible-formats guidance discusses early planning, audience needs and supplier briefs. Its operational context is government communication, so use it as a prompt for the Evidence owner field, not as a private-sector pass certificate.

Six measures, defined

The remaining measures follow the same pattern as the worked row. Each is defined by event; the threshold stays blank for the buyer's authorised owner to set.

For reference, the NCSC publishes Introduction to logging for security purposes, the ICO covers accuracy in its UK GDPR guidance on the accuracy principle, and GOV.UK publishes Accessible communication formats.

Treat failure as designed work

Every standard needs a failure queue. Name who can stop publication, who diagnoses the problem, what evidence is preserved, which controlled copies are affected and who may restart. A missed threshold cannot be cured by changing the denominator after the event.

Separate timeliness from quality. A quick review with a missing rights record fails even if its clock passes. A fully evidenced correction that exceeds a contracted response threshold has one quality result and a different service result. Report both. The buyer should also state how missing timestamps affect the measure rather than excluding inconvenient records silently.

Incident communication needs its own standard set. NCSC guidance for cyber incidents supports preparation and coordination among relevant functions, but supplies no universal clock. Security owns technical facts; legal and regulatory owners decide applicable notifications; communications controls approved wording and channels.

Review a standard when the message scope, affected population, source system, contract, risk or channel changes. Retire it when its event no longer exists, and keep its records as the guide to retiring a channel without losing its records describes. Leave the threshold blank until an authorised owner and a reproducible evidence source are present; otherwise the field cannot support a service claim.

At the review meeting, show the underlying records before any summary. Confirm that the query version matches the data dictionary and that corrections have flowed into past reports. If the measure no longer changes a decision, retire it with a reason. If it still matters but cannot be reproduced, stop reporting it until the evidence chain is repaired.

Before you act

Common questions

How does the worked example row help?

The row shows the level of detail needed for one measure. It gives a numerator, denominator, clock and evidence source. A buyer can copy that structure for other measures, then set its own threshold with an authorised owner.

What is the difference between a quality result and a service result?

The article separates timeliness from quality. A quick review with a missing rights record fails even if its clock passes. A fully evidenced correction that exceeds a contracted response threshold has one quality result and a different service result. Report both.

How should incident communication standards be set?

Incident communication needs its own standard set. NCSC guidance supports preparation and coordination among relevant functions but supplies no universal clock. Security owns technical facts, legal and regulatory owners decide applicable notifications, and communications controls approved wording and channels.